packages/core/src/model-metadata.generated.ts

// Generated by scripts/sync-model-metadata.mjs from https://models.dev/api.json.
// Do not edit by hand; put access-path-specific facts in model-metadata.ts.
import type { ModelMetadata } from './model-metadata.js';

export const GENERATED_MODELS_DEV_METADATA: Record<
  | 'anthropic'
  | 'alibaba'
  | 'alibaba-coding-plan-cn'
  | 'alibaba-coding-plan'
  | 'alibaba-token-plan-cn'
  | 'alibaba-token-plan'
  | 'cerebras'
  | 'cohere'
  | 'cloudflare-workers-ai'
  | 'deepinfra'
  | 'deepseek'
  | 'fireworks-ai'
  | 'github-copilot'
  | 'google'
  | 'gemini-cli'
  | 'groq'
  | 'huggingface'
  | 'kimi-coding-plan'
  | 'MiniMax'
  | 'MiniMax-cn'
  | 'minimax-coding-plan'
  | 'mistral'
  | 'moonshot'
  | 'nvidia'
  | 'ollama-cloud'
  | 'openai'
  | 'opencode'
  | 'opencode-go'
  | 'openrouter'
  | 'siliconflow'
  | 'stepfun'
  | 'stepfun-ai'
  | 'stepfun-ai-step-plan'
  | 'stepfun-step-plan'
  | 'togetherai'
  | 'tencent-coding-plan'
  | 'tencent-token-plan'
  | 'tencent-tokenhub'
  | 'vercel'
  | 'xai'
  | 'xiaomi'
  | 'xiaomi-token-plan-cn'
  | 'xiaomi-token-plan-sgp'
  | 'xiaomi-token-plan-ams'
  | 'zai'
  | 'zai-coding-plan'
  | 'zenmux',
  Record<string, ModelMetadata>
> = {
  anthropic: {
    'claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-haiku-4-5': {
      displayName: 'Claude Haiku 4.5 (latest)',
      description: 'Fast Claude lane for lightweight agents, office tasks, and responsive chat',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      structuredOutput: true,
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-haiku-4-5-20251001': {
      displayName: 'Claude Haiku 4.5',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      structuredOutput: true,
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-5': {
      displayName: 'Claude Opus 4.5 (latest)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-5-20251101': {
      displayName: 'Claude Opus 4.5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2025-11-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-6': {
      displayName: 'Claude Opus 4.6',
      description: 'High-end Claude for difficult coding, planning, and slower expert reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05-31',
      structuredOutput: true,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-7': {
      displayName: 'Claude Opus 4.7',
      description: 'Stronger Opus tier for advanced software work and high-stakes reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-5': {
      displayName: 'Claude Opus 5',
      description: 'Strongest Claude Opus model for coding, agents, and professional work',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-05',
      structuredOutput: true,
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4-5': {
      displayName: 'Claude Sonnet 4.5 (latest)',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-07-31',
      structuredOutput: true,
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4-5-20250929': {
      displayName: 'Claude Sonnet 4.5',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-07-31',
      structuredOutput: true,
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4-6': {
      displayName: 'Claude Sonnet 4.6',
      description:
        'Claude workhorse for coding agents, careful analysis, and production cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://docs.anthropic.com/en/docs/about-claude/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
  },
  alibaba: {
    'qvq-max': {
      displayName: 'QVQ Max',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-03-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen-flash': {
      displayName: 'Qwen Flash',
      description: 'Efficient Qwen model for fast chat, extraction, and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-max': {
      displayName: 'Qwen Max',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 32768,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-mt-plus': {
      displayName: 'Qwen-MT Plus',
      description:
        'Translation model for multilingual conversion, localization, and cross-language workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 16384,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-mt-turbo': {
      displayName: 'Qwen-MT Turbo',
      description:
        'Translation model for multilingual conversion, localization, and cross-language workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 16384,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-omni-turbo': {
      displayName: 'Qwen-Omni Turbo',
      description: 'Qwen omni model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-03-26',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen-omni-turbo-realtime': {
      displayName: 'Qwen-Omni Turbo Realtime',
      description: 'Qwen omni model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-05-08',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen-plus': {
      displayName: 'Qwen Plus',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-09-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-plus-character-ja': {
      displayName: 'Qwen Plus Character (Japanese)',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 8192,
      maxOutputTokens: 512,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-turbo': {
      displayName: 'Qwen Turbo',
      description: 'Efficient Qwen model for fast chat, extraction, and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-04-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-vl-max': {
      displayName: 'Qwen-VL Max',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-08-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen-vl-ocr': {
      displayName: 'Qwen-VL OCR',
      description: 'OCR model for extracting structured text from documents and screenshots',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 34096,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-04-13',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen-vl-plus': {
      displayName: 'Qwen-VL Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-08-15',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen2-5-14b-instruct': {
      displayName: 'Qwen2.5 14B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen2-5-32b-instruct': {
      displayName: 'Qwen2.5 32B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen2-5-72b-instruct': {
      displayName: 'Qwen2.5 72B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen2-5-7b-instruct': {
      displayName: 'Qwen2.5 7B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen2-5-omni-7b': {
      displayName: 'Qwen2.5-Omni 7B',
      description: 'Qwen omni model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen2-5-vl-72b-instruct': {
      displayName: 'Qwen2.5-VL 72B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen2-5-vl-7b-instruct': {
      displayName: 'Qwen2.5-VL 7B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3-14b': {
      displayName: 'Qwen3 14B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-235b-a22b': {
      displayName: 'Qwen3 235B-A22B',
      description: 'Large open Qwen MoE for multilingual reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-32b': {
      displayName: 'Qwen3 32B',
      description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-8b': {
      displayName: 'Qwen3 8B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-asr-flash': {
      displayName: 'Qwen3-ASR Flash',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 53248,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-09-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'qwen3-coder-30b-a3b-instruct': {
      displayName: 'Qwen3-Coder 30B-A3B Instruct',
      description: 'Smaller Qwen coder for efficient local agents and repo-level fixes',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-480b-a35b-instruct': {
      displayName: 'Qwen3-Coder 480B-A35B Instruct',
      description: 'Open Qwen coding heavyweight for repository reasoning and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-flash': {
      displayName: 'Qwen3 Coder Flash',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-plus': {
      displayName: 'Qwen3 Coder Plus',
      description: 'Hosted Qwen coder for software agents, repo edits, and long-context code',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-livetranslate-flash-realtime': {
      displayName: 'Qwen3-LiveTranslate Flash Realtime',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 53248,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-09-22',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen3-max': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen3 model for coding agents, complex reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-next-80b-a3b-instruct': {
      displayName: 'Qwen3-Next 80B-A3B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-next-80b-a3b-thinking': {
      displayName: 'Qwen3-Next 80B-A3B (Thinking)',
      description: 'Efficient Qwen thinking model for local reasoning, math, and coding agents',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-omni-flash': {
      displayName: 'Qwen3-Omni Flash',
      description: 'Qwen omni model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 65536,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-09-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen3-omni-flash-realtime': {
      displayName: 'Qwen3-Omni Flash Realtime',
      description: 'Qwen omni model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 65536,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-09-15',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'qwen3-vl-235b-a22b': {
      displayName: 'Qwen3-VL 235B-A22B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3-vl-30b-a3b': {
      displayName: 'Qwen3-VL 30B-A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3-vl-plus': {
      displayName: 'Qwen3-VL Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.5-122b-a10b': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.5-27b': {
      displayName: 'Qwen3.5 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.5-35b-a3b': {
      displayName: 'Qwen3.5 35B-A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.5-397b-a17b': {
      displayName: 'Qwen3.5 397B-A17B',
      description: 'Large open Qwen multimodal MoE for visual agents and long technical tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-27b': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.6-35b-a3b': {
      displayName: 'Qwen3.6 35B-A3B',
      description: 'Open multimodal Qwen MoE for local agents that need vision, audio, and code',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-max-preview': {
      displayName: 'Qwen3.6 Max Preview',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Earlier Qwen multimodal workhorse for million-token agent and document tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-06-04',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.8-max': {
      displayName: 'Qwen3.8 Max',
      description:
        '2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'qwq-plus': {
      displayName: 'QwQ Plus',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2025-03-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'alibaba-coding-plan-cn': {
    'glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 196608,
      maxOutputTokens: 24576,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-next': {
      displayName: 'Qwen3 Coder Next',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-plus': {
      displayName: 'Qwen3 Coder Plus',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-max-2026-01-23': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-01-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://help.aliyun.com/zh/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'alibaba-coding-plan': {
    'glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 196608,
      inputLimit: 196601,
      maxOutputTokens: 24576,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-next': {
      displayName: 'Qwen3 Coder Next',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder-plus': {
      displayName: 'Qwen3 Coder Plus',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-max-2026-01-23': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-01-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'alibaba-token-plan-cn': {
    'deepseek-v3.2': {
      displayName: 'DeepSeek V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'GLM-5.1',
      description:
        'Strong GLM coding model for agentic engineering, terminals, and repository generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 202752,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'happyhorse-1.1-i2v': {
      displayName: 'HappyHorse 1.1 Image-to-Video',
      description: 'Video model for image-to-video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: [] },
    },
    'happyhorse-1.1-r2v': {
      displayName: 'HappyHorse 1.1 Reference-to-Video',
      description: 'Video model for reference-guided video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: [] },
    },
    'happyhorse-1.1-t2v': {
      displayName: 'HappyHorse 1.1 Text-to-Video',
      description: 'Video model for prompt-driven text-to-video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description:
        'Earlier Kimi frontier model for long-context agents, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Prior MiniMax coding model for agent workflows, office edits, and automation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 196608,
      inputLimit: 196601,
      maxOutputTokens: 32768,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-image-2.0': {
      displayName: 'Qwen Image 2.0',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'qwen-image-2.0-pro': {
      displayName: 'Qwen Image 2.0 Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Earlier Qwen multimodal workhorse for million-token agent and document tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.8-max': {
      displayName: 'Qwen3.8 Max',
      description:
        '2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'qwen3.8-max-preview': {
      displayName: 'Qwen3.8 Max Preview',
      description:
        'Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows',
      lifecycle: 'beta',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'wan2.7-image': {
      displayName: 'Wan2.7 Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'wan2.7-image-pro': {
      displayName: 'Wan2.7 Image Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
  },
  'alibaba-token-plan': {
    'deepseek-v3.2': {
      displayName: 'DeepSeek V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'GLM-5.1',
      description:
        'Strong GLM coding model for agentic engineering, terminals, and repository generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 202752,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'happyhorse-1.1-i2v': {
      displayName: 'HappyHorse 1.1 Image-to-Video',
      description: 'Video model for image-to-video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: [] },
    },
    'happyhorse-1.1-r2v': {
      displayName: 'HappyHorse 1.1 Reference-to-Video',
      description: 'Video model for reference-guided video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: [] },
    },
    'happyhorse-1.1-t2v': {
      displayName: 'HappyHorse 1.1 Text-to-Video',
      description: 'Video model for prompt-driven text-to-video generation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-17',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description:
        'Earlier Kimi frontier model for long-context agents, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Prior MiniMax coding model for agent workflows, office edits, and automation',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 196608,
      inputLimit: 196601,
      maxOutputTokens: 32768,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen-image-2.0': {
      displayName: 'Qwen Image 2.0',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'qwen-image-2.0-pro': {
      displayName: 'Qwen Image 2.0 Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Earlier Qwen multimodal workhorse for million-token agent and document tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.8-max': {
      displayName: 'Qwen3.8 Max',
      description:
        '2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'qwen3.8-max-preview': {
      displayName: 'Qwen3.8 Max Preview',
      description:
        'Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows',
      lifecycle: 'beta',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'wan2.7-image': {
      displayName: 'Wan2.7 Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'wan2.7-image-pro': {
      displayName: 'Wan2.7 Image Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
      contextWindow: 8192,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
  },
  cerebras: {
    'gemma-4-31b': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'beta',
      docsUrl: 'https://inference-docs.cerebras.ai/models/overview',
      contextWindow: 131072,
      maxOutputTokens: 40960,
      structuredOutput: true,
      lastUpdated: '2026-07-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://inference-docs.cerebras.ai/models/overview',
      contextWindow: 131072,
      maxOutputTokens: 40960,
      structuredOutput: true,
      lastUpdated: '2026-06-10',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-glm-4.7': {
      displayName: 'Z.AI GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'beta',
      docsUrl: 'https://inference-docs.cerebras.ai/models/overview',
      contextWindow: 131072,
      maxOutputTokens: 40960,
      structuredOutput: true,
      lastUpdated: '2026-06-10',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none'] },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  cohere: {
    'c4ai-aya-expanse-32b': {
      displayName: 'Aya Expanse 32B',
      description: 'Open multilingual model optimized for generation across 23 languages',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      lastUpdated: '2024-10-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'c4ai-aya-expanse-8b': {
      displayName: 'Aya Expanse 8B',
      description: 'Compact open multilingual model optimized for generation across 23 languages',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 8000,
      maxOutputTokens: 4000,
      lastUpdated: '2024-10-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'c4ai-aya-vision-32b': {
      displayName: 'Aya Vision 32B',
      description:
        'Open multilingual vision model for OCR, visual reasoning, and image question answering',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 16000,
      maxOutputTokens: 4000,
      lastUpdated: '2025-05-14',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'c4ai-aya-vision-8b': {
      displayName: 'Aya Vision 8B',
      description: 'Compact open multilingual vision model for OCR and visual question answering',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 16000,
      maxOutputTokens: 4000,
      lastUpdated: '2025-05-14',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'command-a-03-2025': {
      displayName: 'Command A',
      description: 'Cohere command model for multilingual enterprise agents, tools, and chat',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 256000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2025-03-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-a-plus-05-2026': {
      displayName: 'Command A Plus',
      description:
        "Cohere's stronger command model for multilingual agents and enterprise workflows",
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04-01',
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'command-a-reasoning-08-2025': {
      displayName: 'Command A Reasoning',
      description:
        'Cohere reasoning model for multilingual enterprise agents, tools, and complex workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-a-translate-08-2025': {
      displayName: 'Command A Translate',
      description:
        'Translation model for multilingual conversion, localization, and cross-language workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 8000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2025-08-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-a-vision-07-2025': {
      displayName: 'Command A Vision',
      description:
        'Cohere vision model for multilingual document analysis, OCR, and image understanding',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2025-07-31',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'command-r-08-2024': {
      displayName: 'Command R',
      description: 'Cohere retrieval model for long-context chat and enterprise RAG workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2024-08-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-r-plus-08-2024': {
      displayName: 'Command R+',
      description: "Cohere's RAG workhorse for long-context enterprise search and tool use",
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2024-08-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-r7b-12-2024': {
      displayName: 'Command R7B',
      description: 'Cohere retrieval model for long-context chat and enterprise RAG workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2024-12-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'command-r7b-arabic-02-2025': {
      displayName: 'Command R7B Arabic',
      description:
        'Open Command R model optimized for Arabic enterprise chat, RAG, and cultural knowledge',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      lastUpdated: '2025-02-27',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'north-mini-code-1-0': {
      displayName: 'North Mini Code',
      description: 'Cohere coding model for practical software engineering and agentic edits',
      lifecycle: 'active',
      docsUrl: 'https://docs.cohere.com/docs/models',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-09-23',
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'cloudflare-workers-ai': {
    '@cf/aisingapore/gemma-sea-lion-v4-27b-it': {
      displayName: 'Gemma Sea Lion V4 27B It',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/deepseek-ai/deepseek-r1-distill-qwen-32b': {
      displayName: 'Deepseek R1 Distill Qwen 32B',
      description:
        'Classic open reasoning model for transparent math, coding, and deliberate problem solving',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 80000,
      maxOutputTokens: 80000,
      knowledgeCutoff: '2024-07',
      structuredOutput: false,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/google/gemma-4-26b-a4b-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 256000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '@cf/ibm-granite/granite-4.0-h-micro': {
      displayName: 'Granite 4.0 H Micro',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: false,
      lastUpdated: '2025-10-07',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/meta/llama-3.1-8b-instruct-fp8': {
      displayName: 'Llama 3.1 8B Instruct fp8',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      structuredOutput: false,
      lastUpdated: '2024-07-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/meta/llama-3.2-11b-vision-instruct': {
      displayName: 'Llama 3.2 11B Vision Instruct',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2024-09-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '@cf/meta/llama-3.2-1b-instruct': {
      displayName: 'Llama 3.2 1B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 60000,
      maxOutputTokens: 60000,
      knowledgeCutoff: '2023-12',
      structuredOutput: false,
      lastUpdated: '2024-09-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/meta/llama-3.2-3b-instruct': {
      displayName: 'Llama 3.2 3B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 80000,
      maxOutputTokens: 80000,
      knowledgeCutoff: '2023-12',
      structuredOutput: false,
      lastUpdated: '2024-09-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/meta/llama-3.3-70b-instruct-fp8-fast': {
      displayName: 'Llama 3.3 70B Instruct fp8 Fast',
      description: 'Popular open Llama workhorse for multilingual chat, coding, and self-hosting',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 24000,
      maxOutputTokens: 24000,
      knowledgeCutoff: '2023-12',
      structuredOutput: false,
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/meta/llama-4-scout-17b-16e-instruct': {
      displayName: 'Llama 4 Scout 17B 16E Instruct',
      description: 'Open Llama with long-context vision for efficient multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 131000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08',
      structuredOutput: false,
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '@cf/meta/llama-guard-3-8b': {
      displayName: 'Llama Guard 3 8B',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2025-01-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/mistralai/mistral-small-3.1-24b-instruct': {
      displayName: 'Mistral Small 3.1 24B Instruct',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-03-18',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/moonshotai/kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 262144,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '@cf/moonshotai/kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '@cf/nvidia/nemotron-3-120b-a12b': {
      displayName: 'Nemotron 3 Super 120B',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/qwen/qwen2.5-coder-32b-instruct': {
      displayName: 'Qwen2.5 Coder 32B Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2025-02-27',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/qwen/qwen3-30b-a3b-fp8': {
      displayName: 'Qwen3 30B A3b fp8',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2025-04-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/qwen/qwq-32b': {
      displayName: 'Qwq 32B',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 24000,
      maxOutputTokens: 24000,
      structuredOutput: false,
      lastUpdated: '2025-03-05',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/zai-org/glm-4.7-flash': {
      displayName: 'GLM-4.7-Flash',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    '@cf/zai-org/glm-5.2': {
      displayName: 'Glm 5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://developers.cloudflare.com/workers-ai/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  deepinfra: {
    'deepseek-ai/DeepSeek-R1-0528': {
      displayName: 'DeepSeek-R1-0528',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 163840,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2025-05-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3': {
      displayName: 'DeepSeek-V3',
      description: 'Open DeepSeek MoE chat model for coding, math, and general reasoning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 163840,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2024-12-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.1': {
      displayName: 'DeepSeek-V3.1',
      description: 'Hybrid-reasoning DeepSeek model with thinking and non-thinking modes',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 163840,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.2': {
      displayName: 'DeepSeek-V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 163840,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-26B-A4B-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-4-31B-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-4-E4B-it': {
      displayName: 'Gemma 4 E4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'meta-llama/Llama-3.3-70B-Instruct-Turbo': {
      displayName: 'Llama 3.3 70B Turbo',
      description: 'Compact Llama instruction model for fast chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8': {
      displayName: 'Llama 4 Maverick 17B FP8',
      description: 'Open multimodal Llama model for strong reasoning and fast responses',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta-llama/Llama-4-Scout-17B-16E-Instruct': {
      displayName: 'Llama 4 Scout 17B',
      description: 'Open multimodal Llama model for long-context analysis and efficient agents',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 327680,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.5': {
      displayName: 'MiniMax M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 196608,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description:
        'Open MiniMax flagship for coding agents, office automation, and complex environments',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 196608,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 524288,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.5': {
      displayName: 'Kimi K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.7-Code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/Llama-3.3-Nemotron-Super-49B-v1.5': {
      displayName: 'Llama 3.3 Nemotron Super 49B v1.5',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'deprecated',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/Nemotron-3-Nano-30B-A3B': {
      displayName: 'Nemotron 3 Nano 30B A3B',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2025-12-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning': {
      displayName: 'Nemotron 3 Nano Omni 30B A3B Reasoning',
      description: 'Open Nemotron omni model combining reasoning with text, vision, and audio',
      lifecycle: 'deprecated',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B-Instruct-2507': {
      displayName: 'Qwen3 235B-A22B Instruct 2507',
      description:
        'Updated large open Qwen3 MoE instruct model for multilingual chat, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-07-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-32B': {
      displayName: 'Qwen3 32B',
      description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 40960,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo': {
      displayName: 'Qwen3 Coder 480B A35B Instruct Turbo',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 66536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Max': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen3 model for coding agents, complex reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 256000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Next-80B-A3B-Instruct': {
      displayName: 'Qwen3-Next 80B-A3B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-122B-A10B': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'Qwen/Qwen3.5-27B': {
      displayName: 'Qwen3.5 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'Qwen/Qwen3.5-35B-A3B': {
      displayName: 'Qwen 3.5 35B A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 81920,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-397B-A17B': {
      displayName: 'Qwen 3.5 397B A17B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 81920,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-9B': {
      displayName: 'Qwen3.5 9B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.6-27B': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'Qwen/Qwen3.6-35B-A3B': {
      displayName: 'Qwen3.6 35B A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 81920,
      structuredOutput: true,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.7-Max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 256000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.8-Max': {
      displayName: 'Qwen3.8 Max',
      description:
        '2.4-trillion-parameter MoE flagship for coding, professional work, multimodal understanding, and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 256000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'stepfun-ai/Step-3.7-Flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'tencent/Hy3': {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thinkingmachines/Inkling': {
      displayName: 'Inkling',
      description:
        'Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 524288,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'thinkingmachines/Inkling-Small': {
      displayName: 'Inkling Small',
      description:
        'Multimodal MoE reasoning model (276B total, 12B active) for text, image, and audio',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 524288,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-07-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'XiaomiMiMo/MiMo-V2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'XiaomiMiMo/MiMo-V2.5-Pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'audio'], output: ['text'] },
    },
    'zai-org/GLM-4.6': {
      displayName: 'GLM-4.6',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.7': {
      displayName: 'GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.7-Flash': {
      displayName: 'GLM-4.7-Flash',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-12',
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://deepinfra.com/models',
      contextWindow: 1048576,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  deepseek: {
    'deepseek-chat': {
      displayName: 'DeepSeek Chat',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://api-docs.deepseek.com/quick_start/pricing',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2026-02-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-reasoner': {
      displayName: 'DeepSeek Reasoner',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://api-docs.deepseek.com/quick_start/pricing',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2026-02-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://api-docs.deepseek.com/quick_start/pricing',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://api-docs.deepseek.com/quick_start/pricing',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'fireworks-ai': {
    'accounts/fireworks/models/deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-06-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/deepseek-v4-flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/glm-5p2': {
      displayName: 'GLM 5.2',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1048575,
      maxOutputTokens: 131072,
      lastUpdated: '2026-06-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      lastUpdated: '2026-06-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/kimi-k2p6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/models/kimi-k2p7-code': {
      displayName: 'Kimi K2.7 Code',
      description: 'Kimi coding model for software agents, refactors, and repository reasoning',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      lastUpdated: '2026-06-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/models/kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/models/minimax-m2p7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 196608,
      maxOutputTokens: 196608,
      lastUpdated: '2026-04-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/models/minimax-m3': {
      displayName: 'MiniMax-M3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 512000,
      maxOutputTokens: 512000,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/models/qwen3p7-plus': {
      displayName: 'Qwen 3.7 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/routers/glm-5p2-fast': {
      displayName: 'GLM 5.2 Fast',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1048575,
      maxOutputTokens: 131072,
      lastUpdated: '2026-06-26',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'accounts/fireworks/routers/kimi-k2p6-fast': {
      displayName: 'Kimi K2.6 Fast',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      lastUpdated: '2026-06-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/routers/kimi-k2p6-turbo': {
      displayName: 'Kimi K2.6 Turbo',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/routers/kimi-k2p7-code-fast': {
      displayName: 'Kimi K2.7 Code Fast',
      description: 'Kimi coding model for software agents, refactors, and repository reasoning',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      lastUpdated: '2026-06-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'accounts/fireworks/routers/kimi-k3-fast': {
      displayName: 'Kimi K3 Fast',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://fireworks.ai/docs/',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'github-copilot': {
    'claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-haiku-4.5': {
      displayName: 'Claude Haiku 4.5 (latest)',
      description: 'Fast Claude lane for lightweight agents, office tasks, and responsive chat',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 136000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4.5': {
      displayName: 'Claude Opus 4.5 (latest)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4.6': {
      displayName: 'Claude Opus 4.6',
      description: 'High-end Claude for difficult coding, planning, and slower expert reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-05-31',
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4.7': {
      displayName: 'Claude Opus 4.7',
      description: 'Stronger Opus tier for advanced software work and high-stakes reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4.8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-5': {
      displayName: 'Claude Opus 5',
      description: 'Strongest Claude Opus model for coding, agents, and professional work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1000000,
      inputLimit: 936000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-05',
      structuredOutput: true,
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4': {
      displayName: 'Claude Sonnet 4 (latest)',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 216000,
      inputLimit: 128000,
      maxOutputTokens: 16000,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4.5': {
      displayName: 'Claude Sonnet 4.5 (latest)',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-07-31',
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4.6': {
      displayName: 'Claude Sonnet 4.6',
      description:
        'Claude workhorse for coding agents, careful analysis, and production cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 168000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Reasoning-first Gemini preview for agentic coding and complex problem solving',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1000000,
      inputLimit: 936000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 200000,
      inputLimit: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1000000,
      inputLimit: 936000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-4.1': {
      displayName: 'GPT-4.1',
      description:
        'Long-lived GPT workhorse for coding, instruction following, and production apps',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 128000,
      inputLimit: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5-mini': {
      displayName: 'GPT-5 Mini',
      description: 'Small GPT-5 for responsive agents, coding help, and everyday automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 264000,
      inputLimit: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2': {
      displayName: 'GPT-5.2',
      description:
        'Reliable GPT generation for broad coding, writing, and tool-assisted product work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2-codex': {
      displayName: 'GPT-5.2 Codex',
      description:
        'Code-specialist GPT for repository edits, reviews, and long-running software agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.3-codex': {
      displayName: 'GPT-5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4': {
      displayName: 'GPT-5.4',
      description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4-mini': {
      displayName: 'GPT-5.4 mini',
      description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.4-nano': {
      displayName: 'GPT-5.4 nano',
      description: 'Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.5': {
      displayName: 'GPT-5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna',
      description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-sol': {
      displayName: 'GPT-5.6 Sol',
      description:
        'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-terra': {
      displayName: 'GPT-5.6 Terra',
      description: 'Balanced GPT-5.6 model for capable, cost-efficient everyday work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 500000,
      inputLimit: 372000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 256000,
      inputLimit: 224000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mai-code-1-flash-picker': {
      displayName: 'MAI-Code-1-Flash',
      description:
        'Microsoft coding model built for fast, efficient assistance in everyday developer workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.github.com/en/copilot',
      contextWindow: 256000,
      inputLimit: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12',
      structuredOutput: true,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  google: {
    'deep-research-max-preview-04-2026': {
      displayName: 'Deep Research Max Preview (Apr-21-2026)',
      description:
        'Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] },
    },
    'deep-research-preview-04-2026': {
      displayName: 'Deep Research Preview (Apr-21-2026)',
      description: 'Agentic model for autonomous multi-step research, synthesis, and cited reports',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-2.0-flash': {
      displayName: 'Gemini 2.0 Flash',
      description: 'Earlier Gemini Flash workhorse for responsive multimodal apps and tool use',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2024-12-11',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.0-flash-lite': {
      displayName: 'Gemini 2.0 Flash-Lite',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2024-12-11',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-computer-use-preview-10-2025': {
      displayName: 'Gemini 2.5 Computer Use Preview 10-2025',
      description: 'Specialized Gemini 2.5 model for browser-control agents that automate UI tasks',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-10-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gemini-2.5-flash': {
      displayName: 'Gemini 2.5 Flash',
      description: 'Fast Gemini workhorse for multimodal apps where latency and price matter',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-flash-image': {
      displayName: 'Nano Banana',
      description:
        'Nano Banana image model for fast generation, edits, and character-consistent assets',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2025-08-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-2.5-flash-lite': {
      displayName: 'Gemini 2.5 Flash-Lite',
      description: 'Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-flash-preview-tts': {
      displayName: 'Gemini 2.5 Flash Preview TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-05-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-2.5-pro': {
      displayName: 'Gemini 2.5 Pro',
      description: "Google's proven reasoning model for coding, math, and multimodal analysis",
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-pro-preview-tts': {
      displayName: 'Gemini 2.5 Pro Preview TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-05-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-3-flash-preview': {
      displayName: 'Gemini 3 Flash Preview',
      description:
        'New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3-pro-image': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3-pro-image-preview': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-11-20',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3-pro-preview': {
      displayName: 'Gemini 3 Pro Preview',
      description:
        'Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-11-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-image': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-image-preview': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-lite': {
      displayName: 'Gemini 3.1 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-lite-image': {
      displayName: 'Nano Banana 2 Lite',
      description:
        'Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-lite-preview': {
      displayName: 'Gemini 3.1 Flash Lite Preview',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-live-preview': {
      displayName: 'Gemini 3.1 Flash Live Preview',
      description:
        'High-quality, low-latency Live API model for real-time dialogue and voice-first AI applications',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-03-26',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'gemini-3.1-flash-tts-preview': {
      displayName: 'Gemini 3.1 Flash TTS Preview',
      description: 'Low-latency speech generation with steerable prompts and expressive audio tags',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-15',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Reasoning-first Gemini preview for agentic coding and complex problem solving',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-pro-preview-customtools': {
      displayName: 'Gemini 3.1 Pro Preview Custom Tools',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash-lite': {
      displayName: 'Gemini 3.5 Flash Lite',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-live-translate-preview': {
      displayName: 'Gemini 3.5 Live Translate Preview',
      description:
        'Low-latency audio-to-audio model for real-time speech translation across 70+ languages',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 16384,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-06-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['audio', 'text'] },
    },
    'gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-embedding-001': {
      displayName: 'Gemini Embedding 001',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 2048,
      maxOutputTokens: 1,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gemini-embedding-2': {
      displayName: 'Gemini Embedding 2',
      description:
        'Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 1,
      knowledgeCutoff: '2025-11',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-flash-latest': {
      displayName: 'Gemini Flash Latest',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-flash-lite-latest': {
      displayName: 'Gemini Flash-Lite Latest',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-omni-flash-preview': {
      displayName: 'Gemini Omni Flash Preview',
      description:
        'Video generation and editing model for fast, conversational text- and image-to-video workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'gemini-robotics-er-1.6-preview': {
      displayName: 'Gemini Robotics-ER 1.6 Preview',
      description:
        'Vision-language model for embodied reasoning: spatial understanding, task planning, and physical-world agentic robotics',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'gemma-4-26b-a4b-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gemma-4-31b-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'lyria-3-clip-preview': {
      displayName: 'Lyria 3 Clip Preview',
      description:
        'Music generation model for short 30-second clips, loops, and previews from text or image prompts',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'lyria-3-pro-preview': {
      displayName: 'Lyria 3 Pro Preview',
      description:
        'Music generation model for full-length songs from text or images with vocals and structure',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'veo-3.1-fast-generate-preview': {
      displayName: 'Veo 3.1 fast',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-01-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'veo-3.1-generate-preview': {
      displayName: 'Veo 3.1',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'beta',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'veo-3.1-lite-generate-preview': {
      displayName: 'Veo 3.1 lite',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-31',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
  },
  'gemini-cli': {
    'deep-research-max-preview-04-2026': {
      displayName: 'Deep Research Max Preview (Apr-21-2026)',
      description:
        'Maximum-comprehensiveness agentic researcher for multi-step investigation, synthesis, and cited reports',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] },
    },
    'deep-research-preview-04-2026': {
      displayName: 'Deep Research Preview (Apr-21-2026)',
      description: 'Agentic model for autonomous multi-step research, synthesis, and cited reports',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-2.0-flash': {
      displayName: 'Gemini 2.0 Flash',
      description: 'Earlier Gemini Flash workhorse for responsive multimodal apps and tool use',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2024-12-11',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.0-flash-lite': {
      displayName: 'Gemini 2.0 Flash-Lite',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2024-12-11',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-computer-use-preview-10-2025': {
      displayName: 'Gemini 2.5 Computer Use Preview 10-2025',
      description: 'Specialized Gemini 2.5 model for browser-control agents that automate UI tasks',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-10-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gemini-2.5-flash': {
      displayName: 'Gemini 2.5 Flash',
      description: 'Fast Gemini workhorse for multimodal apps where latency and price matter',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-flash-image': {
      displayName: 'Nano Banana',
      description:
        'Nano Banana image model for fast generation, edits, and character-consistent assets',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2025-08-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-2.5-flash-lite': {
      displayName: 'Gemini 2.5 Flash-Lite',
      description: 'Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-flash-preview-tts': {
      displayName: 'Gemini 2.5 Flash Preview TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-05-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-2.5-pro': {
      displayName: 'Gemini 2.5 Pro',
      description: "Google's proven reasoning model for coding, math, and multimodal analysis",
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-2.5-pro-preview-tts': {
      displayName: 'Gemini 2.5 Pro Preview TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-05-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-3-flash-preview': {
      displayName: 'Gemini 3 Flash Preview',
      description:
        'New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3-pro-image': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3-pro-image-preview': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-11-20',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3-pro-preview': {
      displayName: 'Gemini 3 Pro Preview',
      description:
        'Preview Gemini flagship for complex reasoning, coding, and rich multimodal prompts',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-11-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-image': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-image-preview': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-lite': {
      displayName: 'Gemini 3.1 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-lite-image': {
      displayName: 'Nano Banana 2 Lite',
      description:
        'Fastest, most cost-efficient Gemini image model for high-volume 1K generation and editing',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gemini-3.1-flash-lite-preview': {
      displayName: 'Gemini 3.1 Flash Lite Preview',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-flash-live-preview': {
      displayName: 'Gemini 3.1 Flash Live Preview',
      description:
        'High-quality, low-latency Live API model for real-time dialogue and voice-first AI applications',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-03-26',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio'], output: ['text', 'audio'] },
    },
    'gemini-3.1-flash-tts-preview': {
      displayName: 'Gemini 3.1 Flash TTS Preview',
      description: 'Low-latency speech generation with steerable prompts and expressive audio tags',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-15',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Reasoning-first Gemini preview for agentic coding and complex problem solving',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-pro-preview-customtools': {
      displayName: 'Gemini 3.1 Pro Preview Custom Tools',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash-lite': {
      displayName: 'Gemini 3.5 Flash Lite',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-live-translate-preview': {
      displayName: 'Gemini 3.5 Live Translate Preview',
      description:
        'Low-latency audio-to-audio model for real-time speech translation across 70+ languages',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 16384,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-06-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['audio', 'text'] },
    },
    'gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-embedding-001': {
      displayName: 'Gemini Embedding 001',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 2048,
      maxOutputTokens: 1,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gemini-embedding-2': {
      displayName: 'Gemini Embedding 2',
      description:
        'Multimodal embedding model mapping text, images, video, audio, and PDFs into a unified embedding space',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 1,
      knowledgeCutoff: '2025-11',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-flash-latest': {
      displayName: 'Gemini Flash Latest',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-flash-lite-latest': {
      displayName: 'Gemini Flash-Lite Latest',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-omni-flash-preview': {
      displayName: 'Gemini Omni Flash Preview',
      description:
        'Video generation and editing model for fast, conversational text- and image-to-video workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'gemini-robotics-er-1.6-preview': {
      displayName: 'Gemini Robotics-ER 1.6 Preview',
      description:
        'Vision-language model for embodied reasoning: spatial understanding, task planning, and physical-world agentic robotics',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'gemma-4-26b-a4b-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gemma-4-31b-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'lyria-3-clip-preview': {
      displayName: 'Lyria 3 Clip Preview',
      description:
        'Music generation model for short 30-second clips, loops, and previews from text or image prompts',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'lyria-3-pro-preview': {
      displayName: 'Lyria 3 Pro Preview',
      description:
        'Music generation model for full-length songs from text or images with vocals and structure',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'veo-3.1-fast-generate-preview': {
      displayName: 'Veo 3.1 fast',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-01-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'veo-3.1-generate-preview': {
      displayName: 'Veo 3.1',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'beta',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'veo-3.1-lite-generate-preview': {
      displayName: 'Veo 3.1 lite',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://ai.google.dev/gemini-api/docs/models',
      contextWindow: 480,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-31',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
  },
  groq: {
    'allam-2-7b': {
      displayName: 'ALLaM-2-7b',
      description: 'ALLaM-2-7b instruction tuned model by SDAIA',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 4096,
      maxOutputTokens: 4096,
      lastUpdated: '2025-01-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'canopylabs/orpheus-arabic-saudi': {
      displayName: 'Canopy Labs Orpheus Arabic Saudi',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'beta',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 4000,
      maxOutputTokens: 50000,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'canopylabs/orpheus-v1-english': {
      displayName: 'Canopy Labs Orpheus V1 English',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'beta',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 4000,
      maxOutputTokens: 50000,
      lastUpdated: '2025-12-19',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'groq/compound': {
      displayName: 'Compound',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      lastUpdated: '2025-09-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'groq/compound-mini': {
      displayName: 'Compound Mini',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      lastUpdated: '2025-09-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'llama-3.1-8b-instant': {
      displayName: 'Llama 3.1 8B',
      description: 'Compact Llama instruction model for fast chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'llama-3.3-70b-versatile': {
      displayName: 'Llama 3.3 70B',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-prompt-guard-2-22m': {
      displayName: 'Llama Prompt Guard 2 22M',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'beta',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 512,
      maxOutputTokens: 512,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-prompt-guard-2-86m': {
      displayName: 'Prompt Guard 2 86M',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'beta',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 512,
      maxOutputTokens: 512,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2025-10-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2025-09-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-safeguard-20b': {
      displayName: 'Safety GPT OSS 20B',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'beta',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-06-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.6-27b': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'default'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'whisper-large-v3': {
      displayName: 'Whisper',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'whisper-large-v3-turbo': {
      displayName: 'Whisper Large V3 Turbo',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://console.groq.com/docs/models',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2024-10-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
  },
  huggingface: {
    'deepseek-ai/DeepSeek-R1': {
      displayName: 'DeepSeek-R1',
      description:
        'Classic open reasoning model for transparent math, coding, and deliberate problem solving',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 64000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-R1-0528': {
      displayName: 'DeepSeek-R1-0528',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 163840,
      maxOutputTokens: 163840,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3': {
      displayName: 'DeepSeek-V3',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 64000,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2024-12-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.1': {
      displayName: 'DeepSeek-V3.1',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.2': {
      displayName: 'DeepSeek-V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 163840,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-12-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1048576,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1048576,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1048576,
      maxOutputTokens: 393216,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-26B-A4B-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-4-31B-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta-llama/Llama-3.3-70B-Instruct': {
      displayName: 'Llama-3.3-70B-Instruct',
      description: 'Popular open Llama workhorse for multilingual chat, coding, and self-hosting',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: true,
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2': {
      displayName: 'MiniMax-M2',
      description: 'Efficient open MiniMax model built for coding agents and tool-heavy workflows',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 128000,
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-10',
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 524288,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2-Instruct': {
      displayName: 'Kimi-K2-Instruct',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-07-14',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/Kimi-K2-Instruct-0905': {
      displayName: 'Kimi-K2-Instruct-0905',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/Kimi-K2-Thinking': {
      displayName: 'Kimi-K2-Thinking',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.5': {
      displayName: 'Kimi-K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-01-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.6': {
      displayName: 'Kimi-K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.7-Code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K3': {
      displayName: 'Kimi K3',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B': {
      displayName: 'Qwen3 235B-A22B',
      description: 'Large open Qwen MoE for multilingual reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 40960,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B-Instruct-2507': {
      displayName: 'Qwen3 235B-A22B Instruct 2507',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2025-07-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B-Thinking-2507': {
      displayName: 'Qwen3-235B-A22B-Thinking-2507',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-32B': {
      displayName: 'Qwen3 32B',
      description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-30B-A3B-Instruct': {
      displayName: 'Qwen3-Coder 30B-A3B Instruct',
      description: 'Smaller Qwen coder for efficient local agents and repo-level fixes',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-480B-A35B-Instruct': {
      displayName: 'Qwen3-Coder-480B-A35B-Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 66536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-Next': {
      displayName: 'Qwen3-Coder-Next',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Embedding-4B': {
      displayName: 'Qwen 3 Embedding 4B',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 32000,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-01-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Embedding-8B': {
      displayName: 'Qwen 3 Embedding 8B',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 32000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-01-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Next-80B-A3B-Instruct': {
      displayName: 'Qwen3-Next-80B-A3B-Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 66536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-11',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Next-80B-A3B-Thinking': {
      displayName: 'Qwen3-Next-80B-A3B-Thinking',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-11',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-122B-A10B': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-27B': {
      displayName: 'Qwen3.5 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-35B-A3B': {
      displayName: 'Qwen3.5 35B-A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-397B-A17B': {
      displayName: 'Qwen3.5-397B-A17B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-9B': {
      displayName: 'Qwen3.5 9B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.6-27B': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.6-35B-A3B': {
      displayName: 'Qwen3.6 35B-A3B',
      description: 'Open multimodal Qwen MoE for local agents that need vision, audio, and code',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'stepfun-ai/Step-3.5-Flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun-ai/Step-3.7-Flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'tencent/Hy3': {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thinkingmachines/Inkling': {
      displayName: 'Inkling',
      description: 'Multimodal model for analyzing text, images, documents, and rich media',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'thinkingmachines/Inkling-Small': {
      displayName: 'Inkling Small',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 524288,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-07-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'XiaomiMiMo/MiMo-V2-Flash': {
      displayName: 'MiMo-V2-Flash',
      description: 'MiMo flash model for fast multimodal assistance and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'XiaomiMiMo/MiMo-V2.5': {
      displayName: 'MiMo-V2.5',
      description: 'MiMo model for long-context reasoning, perception, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'XiaomiMiMo/MiMo-V2.5-Pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.5': {
      displayName: 'GLM-4.5',
      description: 'Hybrid-reasoning GLM release that made the 4.5 line broadly useful',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.5-Air': {
      displayName: 'GLM-4.5-Air',
      description: 'Lighter GLM-4.5 variant for fast coding assistance and cheaper agents',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.5V': {
      displayName: 'GLM-4.5V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 65536,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-08-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'zai-org/GLM-4.6': {
      displayName: 'GLM-4.6',
      description: 'Late GLM-4 workhorse for coding agents, reasoning, and structured tasks',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.7': {
      displayName: 'GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.7-Flash': {
      displayName: 'GLM-4.7-Flash',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-08-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-03',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://huggingface.co/docs/inference-providers',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'kimi-coding-plan': {
    k3: {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://www.kimi.com/code/docs/en/third-party-tools/other-coding-agents.html',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'k3-256k': {
      displayName: 'Kimi K3-256K',
      description:
        '256K-context version of Kimi K3, reducing token consumption for shorter coding sessions',
      lifecycle: 'active',
      docsUrl: 'https://www.kimi.com/code/docs/en/third-party-tools/other-coding-agents.html',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-for-coding': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://www.kimi.com/code/docs/en/third-party-tools/other-coding-agents.html',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-for-coding-highspeed': {
      displayName: 'Kimi For Coding HighSpeed',
      description: 'Lower-latency Kimi Code variant for interactive edits and coding-agent loops',
      lifecycle: 'active',
      docsUrl: 'https://www.kimi.com/code/docs/en/third-party-tools/other-coding-agents.html',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  MiniMax: {
    'MiniMax-M2': {
      displayName: 'MiniMax-M2',
      description: 'Efficient open MiniMax model built for coding agents and tool-heavy workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 196608,
      maxOutputTokens: 128000,
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'Earlier MiniMax agent model for practical coding and productivity tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Prior MiniMax coding model for agent workflows, office edits, and automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5-highspeed': {
      displayName: 'MiniMax-M2.5-highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description:
        'Open MiniMax flagship for coding agents, office automation, and complex environments',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7-highspeed': {
      displayName: 'MiniMax-M2.7-highspeed',
      description: 'Low-latency M2.7 variant for interactive coding plans and agent loops',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/guides/quickstart',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'MiniMax-cn': {
    'MiniMax-M2': {
      displayName: 'MiniMax-M2',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 196608,
      maxOutputTokens: 128000,
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5-highspeed': {
      displayName: 'MiniMax-M2.5-highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7-highspeed': {
      displayName: 'MiniMax-M2.7-highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimaxi.com/docs/guides/quickstart',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'minimax-coding-plan': {
    'MiniMax-M2': {
      displayName: 'MiniMax-M2',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 196608,
      maxOutputTokens: 128000,
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.5-highspeed': {
      displayName: 'MiniMax-M2.5-highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M2.7-highspeed': {
      displayName: 'MiniMax-M2.7-highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.minimax.io/docs/token-plan/intro',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  mistral: {
    'codestral-latest': {
      displayName: 'Codestral (latest)',
      description: 'Mistral code model for completions, refactors, and developer IDE workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 256000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-01-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-2512': {
      displayName: 'Devstral 2',
      description:
        "Mistral's coding-agent model for repository work, terminal tasks, and software fixes",
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-12',
      lastUpdated: '2025-12-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-latest': {
      displayName: 'Devstral 2',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-12',
      lastUpdated: '2025-12-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-medium-2507': {
      displayName: 'Devstral Medium',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-07-10',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-medium-latest': {
      displayName: 'Devstral 2 (latest)',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-12',
      lastUpdated: '2025-12-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-small-2505': {
      displayName: 'Devstral Small 2505',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-07',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'devstral-small-2507': {
      displayName: 'Devstral Small',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-07-10',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'labs-devstral-small-2512': {
      displayName: 'Devstral Small 2',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-12',
      lastUpdated: '2025-12-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'magistral-medium-latest': {
      displayName: 'Magistral Medium (latest)',
      description: 'Mistral reasoning model for transparent analysis, math, and complex decisions',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2025-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'magistral-small': {
      displayName: 'Magistral Small',
      description: 'Mistral reasoning model for transparent analysis, math, and complex decisions',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2025-03-17',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ministral-3b-latest': {
      displayName: 'Ministral 3B (latest)',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-10-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ministral-8b-latest': {
      displayName: 'Ministral 8B (latest)',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-10-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral-embed': {
      displayName: 'Mistral Embed',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 8000,
      maxOutputTokens: 3072,
      lastUpdated: '2023-12-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral-large-2411': {
      displayName: 'Mistral Large 2.1',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-11',
      lastUpdated: '2024-11-18',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral-large-2512': {
      displayName: 'Mistral Large 3',
      description:
        "Mistral's largest general model for enterprise agents, coding, and multilingual reasoning",
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-11',
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-large-latest': {
      displayName: 'Mistral Large (latest)',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-11',
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-medium-2505': {
      displayName: 'Mistral Medium 3',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-07',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-medium-2508': {
      displayName: 'Mistral Medium 3.1',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-08-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-medium-2604': {
      displayName: 'Mistral Medium 3.5',
      description: 'Balanced Mistral model for enterprise assistants, multilingual work, and tools',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-medium-latest': {
      displayName: 'Mistral Medium (latest)',
      description: 'Balanced Mistral model for enterprise assistants, multilingual work, and tools',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-nemo': {
      displayName: 'Mistral Nemo',
      description: 'Efficient Mistral-NVIDIA open model for multilingual chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2024-07-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral-small-2506': {
      displayName: 'Mistral Small 3.2',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-03',
      lastUpdated: '2025-06-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-small-2603': {
      displayName: 'Mistral Small 4',
      description: 'Fast Mistral production model for chat, extraction, and cost-sensitive agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-03-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-small-latest': {
      displayName: 'Mistral Small (latest)',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-03-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'open-mistral-7b': {
      displayName: 'Mistral 7B',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 8000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2023-09-27',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'open-mistral-nemo': {
      displayName: 'Open Mistral Nemo',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2024-07-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'open-mixtral-8x22b': {
      displayName: 'Mixtral 8x22B',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 64000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-04-17',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'open-mixtral-8x7b': {
      displayName: 'Mixtral 8x7B',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-01',
      lastUpdated: '2023-12-11',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'pixtral-12b': {
      displayName: 'Pixtral 12B',
      description: 'Mistral vision-language model for image understanding and multimodal chat',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09',
      lastUpdated: '2024-09-01',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'pixtral-large-latest': {
      displayName: 'Pixtral Large (latest)',
      description: "Mistral's larger vision model for document-heavy image understanding and chat",
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-11',
      lastUpdated: '2024-11-04',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'voxtral-mini-latest': {
      displayName: 'Voxtral Mini (latest)',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'voxtral-mini-tts-latest': {
      displayName: 'Voxtral Mini TTS (latest)',
      description: 'Multilingual text-to-speech model with zero-shot voice cloning',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'voxtral-small-latest': {
      displayName: 'Voxtral Small (latest)',
      description: 'Instruct model with native audio input for speech understanding and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.mistral.ai/getting-started/models/',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2025-07-15',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'audio'], output: ['text'] },
    },
  },
  moonshot: {
    'kimi-k2-0711-preview': {
      displayName: 'Kimi K2 0711',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-07-14',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2-0905-preview': {
      displayName: 'Kimi K2 0905',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2-thinking': {
      displayName: 'Kimi K2 Thinking',
      description:
        'Thinking Kimi model for slower research passes, planning, and hard technical questions',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2-thinking-turbo': {
      displayName: 'Kimi K2 Thinking Turbo',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2-turbo-preview': {
      displayName: 'Kimi K2 Turbo',
      description: 'Fast Kimi model for responsive chat, coding help, and agent loops',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description:
        'Earlier Kimi frontier model for long-context agents, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code-highspeed': {
      displayName: 'Kimi K2.7 Code HighSpeed',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://platform.moonshot.cn/docs/api/chat',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  nvidia: {
    'abacusai/dracarys-llama-3.1-70b-instruct': {
      displayName: 'dracarys-llama-3.1-70b-instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-05-22',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'baai/bge-m3': {
      displayName: 'BGE M3',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 8192,
      maxOutputTokens: 1024,
      lastUpdated: '2026-04-30',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'black-forest-labs/flux_1-kontext-dev': {
      displayName: 'FLUX.1-Kontext-dev',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 40960,
      maxOutputTokens: 40960,
      lastUpdated: '2025-08-12',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['image'] },
    },
    'black-forest-labs/flux_1-schnell': {
      displayName: 'FLUX.1-schnell',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 77,
      inputLimit: 77,
      maxOutputTokens: 0,
      knowledgeCutoff: '2024-07',
      structuredOutput: false,
      lastUpdated: '2026-02-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'black-forest-labs/flux_2-klein-4b': {
      displayName: 'FLUX.2 Klein 4B',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 40960,
      maxOutputTokens: 40960,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-01-31',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: ['image'] },
    },
    'black-forest-labs/flux.1-dev': {
      displayName: 'FLUX.1-dev',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 4096,
      maxOutputTokens: 0,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bytedance/seed-oss-36b-instruct': {
      displayName: 'ByteDance-Seed/Seed-OSS-36B-Instruct',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1048576,
      maxOutputTokens: 393216,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1048576,
      maxOutputTokens: 393216,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-2-2b-it': {
      displayName: 'Gemma 2 2b It',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      structuredOutput: true,
      lastUpdated: '2024-07-16',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-3-12b-it': {
      displayName: 'Gemma 3 12B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2025-03-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3-4b-it': {
      displayName: 'Gemma 3 4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2025-03-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3n-e2b-it': {
      displayName: 'Gemma 3n E2b It',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2025-06-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3n-e4b-it': {
      displayName: 'Gemma 3n E4b It',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2025-06-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-4-31b-it': {
      displayName: 'Gemma-4-31B-IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 256000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/google-paligemma': {
      displayName: 'paligemma',
      description: 'Gemini multimodal model for text, image, audio, video, and document tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2024-08-26',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/esm2-650m': {
      displayName: 'esm2-650m',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-03-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/esmfold': {
      displayName: 'esmfold',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-06-12',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.1-70b-instruct': {
      displayName: 'Llama 3.1 70b Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      structuredOutput: true,
      lastUpdated: '2024-07-16',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.1-8b-instruct': {
      displayName: 'Llama 3.1 8B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 16000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2025-01-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.2-11b-vision-instruct': {
      displayName: 'Llama 3.2 11b Vision Instruct',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: true,
      lastUpdated: '2024-09-18',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/llama-3.2-1b-instruct': {
      displayName: 'Llama 3.2 1b Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: true,
      lastUpdated: '2024-09-18',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.2-3b-instruct': {
      displayName: 'Llama 3.2 3B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 32000,
      structuredOutput: true,
      lastUpdated: '2024-09-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.2-90b-vision-instruct': {
      displayName: 'Llama-3.2-90B-Vision-Instruct',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-09-25',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/llama-3.3-70b-instruct': {
      displayName: 'Llama 3.3 70b Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      structuredOutput: true,
      lastUpdated: '2024-11-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-4-maverick-17b-128e-instruct': {
      displayName: 'Llama 4 Maverick 17b 128e Instruct',
      description: 'Open multimodal Llama model for strong reasoning and fast responses',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-02',
      structuredOutput: true,
      lastUpdated: '2025-04-01',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/llama-guard-4-12b': {
      displayName: 'Llama Guard 4 12B',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      lastUpdated: '2026-04-30',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'microsoft/phi-4-mini-instruct': {
      displayName: 'Phi-4-Mini',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'microsoft/phi-4-multimodal-instruct': {
      displayName: 'Phi 4 Multimodal',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      inputLimit: 128000,
      maxOutputTokens: 16384,
      structuredOutput: false,
      lastUpdated: '2025-07-26',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimaxai/minimax-m2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimaxai/minimax-m3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1000000,
      maxOutputTokens: 16384,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/magistral-small-2506': {
      displayName: 'Magistral Small 2506',
      description: 'Mistral reasoning model for transparent analysis, math, and complex decisions',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      inputLimit: 32768,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2025-09-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/ministral-14b-instruct-2512': {
      displayName: 'Ministral 3 14B Instruct 2512',
      description: 'Compact Mistral VLM for chat and instruction-based workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-7b-instruct-v0.3': {
      displayName: 'Mistral-7B-Instruct-v0.3',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2025-04-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/mistral-large-3-675b-instruct-2512': {
      displayName: 'Mistral Large 3 675B Instruct 2512',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-medium-3-instruct': {
      displayName: 'Mistral Medium 3',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      inputLimit: 131072,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2025-09-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-medium-3.5-128b': {
      displayName: 'Mistral Medium 3.5',
      description: 'Balanced Mistral model for enterprise assistants, multilingual work, and tools',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-nemotron': {
      displayName: 'mistral-nemotron',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-06-12',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/mistral-small-4-119b-2603': {
      displayName: 'mistral-small-4-119b-2603',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2026-03-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mixtral-8x22b-instruct': {
      displayName: 'Mistral: Mixtral 8x22B Instruct',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 65536,
      maxOutputTokens: 13108,
      lastUpdated: '2024-04-17',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/mixtral-8x7b-instruct': {
      displayName: 'Mistral: Mixtral 8x7B Instruct',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-15',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-instruct-0905': {
      displayName: 'Kimi K2 0905',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/active-speaker-detection': {
      displayName: 'Active Speaker Detection',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2026-04-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: [], output: ['text'] },
    },
    'nvidia/bevformer': {
      displayName: 'bevformer',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-07-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: [], output: ['text'] },
    },
    'nvidia/cosmos-predict1-5b': {
      displayName: 'cosmos-predict1-5b',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2025-03-18',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'nvidia/cosmos-reason2-8b': {
      displayName: 'Cosmos Reason2 8B',
      description:
        'Vision language model for physical-world understanding with structured reasoning on video and images',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2025-12-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/cosmos-transfer1-7b': {
      displayName: 'cosmos-transfer1-7b',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2025-06-30',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'nvidia/cosmos-transfer2_5-2b': {
      displayName: 'cosmos-transfer2.5-2b',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'nvidia/gliner-pii': {
      displayName: 'gliner-pii',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2026-03-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3_2-nemoretriever-300m-embed-v1': {
      displayName: 'llama-3_2-nemoretriever-300m-embed-v1',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      lastUpdated: '2025-07-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.1-nemotron-70b-instruct': {
      displayName: 'Llama 3.1 Nemotron 70B Instruct',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-04-15',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.1-nemotron-nano-8b-v1': {
      displayName: 'Llama 3.1 Nemotron Nano 8B v1',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2025-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.1-nemotron-nano-vl-8b-v1': {
      displayName: 'Llama 3.1 Nemotron Nano VL 8B v1',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 16384,
      lastUpdated: '2025-04-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/llama-3.1-nemotron-safety-guard-8b-v3': {
      displayName: 'llama-3.1-nemotron-safety-guard-8b-v3',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2025-10-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.1-nemotron-ultra-253b-v1': {
      displayName: 'Llama 3.1 Nemotron Ultra 253B',
      description: 'Flagship Nemotron model for high-throughput reasoning and complex agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      lastUpdated: '2025-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.3-nemotron-super-49b-v1': {
      displayName: 'Llama 3.3 Nemotron Super 49B v1',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      lastUpdated: '2025-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-3.3-nemotron-super-49b-v1.5': {
      displayName: 'Llama 3.3 Nemotron Super 49B v1.5',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/llama-nemotron-embed-vl-1b-v2': {
      displayName: 'llama-nemotron-embed-vl-1b-v2',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      lastUpdated: '2026-02-10',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/llama-nemotron-rerank-vl-1b-v2': {
      displayName: 'llama-nemotron-rerank-vl-1b-v2',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2026-03-31',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/magpie-tts-zeroshot': {
      displayName: 'magpie-tts-zeroshot',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2025-06-12',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['audio'] },
    },
    'nvidia/nemotron-3-content-safety': {
      displayName: 'nemotron-3-content-safety',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2026-04-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-30b-a3b': {
      displayName: 'nemotron-3-nano-30b-a3b',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-09',
      lastUpdated: '2024-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning': {
      displayName: 'Nemotron 3 Nano Omni',
      description: 'Open Nemotron omni model combining reasoning with text, vision, and audio',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 256000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'nvidia/nemotron-3-super-120b-a12b': {
      displayName: 'Nemotron 3 Super',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-ultra-550b-a55b': {
      displayName: 'Nemotron 3 Ultra 550B A55B',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-content-safety-reasoning-4b': {
      displayName: 'nemotron-content-safety-reasoning-4b',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2026-01-22',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-mini-4b-instruct': {
      displayName: 'nemotron-mini-4b-instruct',
      description: 'Compact Nemotron model for efficient reasoning and deployable AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2024-08-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-nano-12b-v2-vl': {
      displayName: 'Nemotron Nano 12B v2 VL',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      lastUpdated: '2025-10-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/nemotron-voicechat': {
      displayName: 'nemotron-voicechat',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'audio'], output: ['text'] },
    },
    'nvidia/nv-embed-v1': {
      displayName: 'nv-embed-v1',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      lastUpdated: '2025-07-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nv-embedcode-7b-v1': {
      displayName: 'nv-embedcode-7b-v1',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 32768,
      maxOutputTokens: 2048,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nvidia-nemotron-nano-9b-v2': {
      displayName: 'nvidia-nemotron-nano-9b-v2',
      description: 'Compact Nemotron model for efficient reasoning and deployable AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-09',
      lastUpdated: '2025-08-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/rerank-qa-mistral-4b': {
      displayName: 'rerank-qa-mistral-4b',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2025-01-17',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/riva-translate-4b-instruct-v1.1': {
      displayName: 'riva-translate-4b-instruct-v1_1',
      description:
        'Translation model for multilingual conversion, localization, and cross-language workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2025-12-12',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/sparsedrive': {
      displayName: 'sparsedrive',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-07-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: [], output: ['text'] },
    },
    'nvidia/streampetr': {
      displayName: 'streampetr',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-11-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: [], output: ['text'] },
    },
    'nvidia/studiovoice': {
      displayName: 'studiovoice',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-06-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/synthetic-video-detector': {
      displayName: 'synthetic-video-detector',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2026-04-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: [], output: ['text'] },
    },
    'nvidia/usdcode': {
      displayName: 'usdcode',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      lastUpdated: '2026-01-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/usdvalidate': {
      displayName: 'usdvalidate',
      description: 'Nemotron model for efficient reasoning, coding, and specialized AI agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      lastUpdated: '2025-01-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT-OSS-120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-08',
      structuredOutput: true,
      lastUpdated: '2025-08-14',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/whisper-large-v3': {
      displayName: 'Whisper Large v3',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-09',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'poolside/laguna-xs-2.1': {
      displayName: 'Laguna XS 2.1',
      description: 'Agentic coding model from Poolside in the XS size class for local deployment',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      structuredOutput: false,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-image': {
      displayName: 'Qwen Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 0,
      structuredOutput: false,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['image'] },
    },
    'qwen/qwen-image-edit': {
      displayName: 'Qwen Image Edit',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 0,
      maxOutputTokens: 0,
      structuredOutput: false,
      lastUpdated: '2025-08-19',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['image'] },
    },
    'qwen/qwen2.5-coder-32b-instruct': {
      displayName: 'Qwen2.5 Coder 32b Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      structuredOutput: true,
      lastUpdated: '2024-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder-480b-a35b-instruct': {
      displayName: 'Qwen3 Coder 480B A35B Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 66536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-next-80b-a3b-instruct': {
      displayName: 'Qwen3-Next-80B-A3B-Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.5-122b-a10b': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'qwen/qwen3.5-397b-a17b': {
      displayName: 'Qwen3.5-397B-A17B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 262144,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2026-01',
      structuredOutput: true,
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'sarvamai/sarvam-m': {
      displayName: 'sarvam-m',
      description:
        'Efficient Indian-language reasoning model for chat, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun-ai/step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 256000,
      maxOutputTokens: 16384,
      lastUpdated: '2026-02-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun-ai/step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 256000,
      maxOutputTokens: 16384,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'thinkingmachines/inkling': {
      displayName: 'Inkling',
      description:
        'Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'upstage/solar-10.7b-instruct': {
      displayName: 'solar-10.7b-instruct',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      lastUpdated: '2025-04-10',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.api.nvidia.com/nim/',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'ollama-cloud': {
    'deepseek-v4-flash': {
      displayName: 'deepseek-v4-flash',
      description: 'Fast DeepSeek model for efficient chat, coding help, and agent loops',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash:0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'deepseek-v4-pro',
      description: 'Flagship DeepSeek model for coding, reasoning, and agentic work',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gemma4:31b': {
      displayName: 'gemma4:31b',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'glm-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 976000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-oss:120b': {
      displayName: 'gpt-oss:120b',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-oss:20b': {
      displayName: 'gpt-oss:20b',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'kimi-k2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'kimi-k2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2026-04-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'kimi-k2.7-code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k3': {
      displayName: 'kimi-k3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'minimax-m2.5': {
      displayName: 'minimax-m2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.7': {
      displayName: 'minimax-m2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 196608,
      maxOutputTokens: 196608,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m3': {
      displayName: 'minimax-m3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 512000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-31',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral-large-3:675b': {
      displayName: 'mistral-large-3:675b',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2026-01-19',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nemotron-3-nano:30b': {
      displayName: 'nemotron-3-nano:30b',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nemotron-3-super': {
      displayName: 'nemotron-3-super',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      lastUpdated: '2026-03-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nemotron-3-ultra': {
      displayName: 'nemotron-3-ultra',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.5:397b': {
      displayName: 'qwen3.5:397b',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.ollama.com/cloud',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      lastUpdated: '2026-02-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  openai: {
    'chatgpt-image-latest': {
      displayName: 'chatgpt-image-latest',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 0,
      inputLimit: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gpt-3.5-turbo': {
      displayName: 'GPT-3.5-turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 16385,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09-01',
      structuredOutput: false,
      lastUpdated: '2023-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-4': {
      displayName: 'GPT-4',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2023-11',
      structuredOutput: false,
      lastUpdated: '2024-04-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-4-turbo': {
      displayName: 'GPT-4 Turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: false,
      lastUpdated: '2024-04-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-4.1': {
      displayName: 'GPT-4.1',
      description:
        'Long-lived GPT workhorse for coding, instruction following, and production apps',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-4.1-mini': {
      displayName: 'GPT-4.1 mini',
      description: 'Affordable GPT-4.1 lane for fast coding help and structured extraction',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-4.1-nano': {
      displayName: 'GPT-4.1 nano',
      description: 'Tiny GPT-4.1 option for classification, routing, and very high-volume tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-4o': {
      displayName: 'GPT-4o',
      description: 'Omni-era GPT for multimodal chat, practical coding, and general assistants',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-08-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-4o-2024-05-13': {
      displayName: 'GPT-4o (2024-05-13)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-05-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-4o-2024-08-06': {
      displayName: 'GPT-4o (2024-08-06)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-08-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-4o-2024-11-20': {
      displayName: 'GPT-4o (2024-11-20)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-11-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-4o-mini': {
      displayName: 'GPT-4o mini',
      description: 'Small omni GPT for cheap multimodal assistance and production-scale traffic',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-07-18',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5': {
      displayName: 'GPT-5',
      description: 'Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5-mini': {
      displayName: 'GPT-5 Mini',
      description: 'Small GPT-5 for responsive agents, coding help, and everyday automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5-nano': {
      displayName: 'GPT-5 Nano',
      description: 'Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5-pro': {
      displayName: 'GPT-5 Pro',
      description: 'Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 272000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-10-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.1': {
      displayName: 'GPT-5.1',
      description: 'Sharper GPT-5 generation for coding, product work, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2': {
      displayName: 'GPT-5.2',
      description:
        'Reliable GPT generation for broad coding, writing, and tool-assisted product work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2-chat-latest': {
      displayName: 'GPT-5.2 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2-pro': {
      displayName: 'GPT-5.2 Pro',
      description: 'Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: false,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.3-chat-latest': {
      displayName: 'GPT-5.3 Chat (latest)',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.3-codex': {
      displayName: 'GPT-5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.3-codex-spark': {
      displayName: 'GPT-5.3 Codex Spark',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      inputLimit: 100000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4': {
      displayName: 'GPT-5.4',
      description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4-mini': {
      displayName: 'GPT-5.4 mini',
      description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.4-nano': {
      displayName: 'GPT-5.4 nano',
      description: 'Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.4-pro': {
      displayName: 'GPT-5.4 Pro',
      description: 'More exact GPT-5.4 tier for demanding professional reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: false,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.5': {
      displayName: 'GPT-5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.5-pro': {
      displayName: 'GPT-5.5 Pro',
      description: 'Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6': {
      displayName: 'GPT-5.6',
      description:
        'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna',
      description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-sol': {
      displayName: 'GPT-5.6 Sol',
      description:
        'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-terra': {
      displayName: 'GPT-5.6 Terra',
      description: 'Balanced GPT-5.6 model for capable, cost-efficient everyday work',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-image-1': {
      displayName: 'gpt-image-1',
      description:
        'OpenAI image model for production generation, edits, and brand-safe visual workflows',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 0,
      inputLimit: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-04-24',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['image'] },
    },
    'gpt-image-1-mini': {
      displayName: 'gpt-image-1-mini',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 0,
      inputLimit: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-09-26',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gpt-image-1.5': {
      displayName: 'gpt-image-1.5',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 0,
      inputLimit: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'gpt-image-2': {
      displayName: 'gpt-image-2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 0,
      inputLimit: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['image'] },
    },
    'gpt-realtime-2.1': {
      displayName: 'GPT-Realtime-2.1',
      description:
        'Realtime speech-to-speech model with configurable reasoning, tool use, and robust voice-agent behavior',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 128000,
      inputLimit: 96000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: false,
      lastUpdated: '2026-07-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'audio', 'image'], output: ['text', 'audio'] },
    },
    o1: {
      displayName: 'o1',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-12-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'o1-pro': {
      displayName: 'o1-pro',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2025-03-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    o3: {
      displayName: 'o3',
      description: 'Deliberate o-series reasoner for hard math, coding, and multi-step analysis',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'o3-mini': {
      displayName: 'o3-mini',
      description: 'Smaller o-series reasoner for economical coding, math, and planning tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-01-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'o3-pro': {
      displayName: 'o3-pro',
      description: 'High-effort o3 tier for difficult technical reasoning and careful answers',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-06-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'o4-mini': {
      displayName: 'o4-mini',
      description: 'Fast o-series model for compact reasoning, coding, and tool use',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'text-embedding-3-large': {
      displayName: 'text-embedding-3-large',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 8191,
      maxOutputTokens: 3072,
      knowledgeCutoff: '2024-01',
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'text-embedding-3-small': {
      displayName: 'text-embedding-3-small',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 8191,
      maxOutputTokens: 1536,
      knowledgeCutoff: '2024-01',
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'text-embedding-ada-002': {
      displayName: 'text-embedding-ada-002',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl: 'https://platform.openai.com/docs/models',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      knowledgeCutoff: '2022-12',
      lastUpdated: '2022-12-15',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  opencode: {
    'big-pickle': {
      displayName: 'Big Pickle',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      inputLimit: 160000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-10-17',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'claude-3-5-haiku': {
      displayName: 'Claude Haiku 3.5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-07-31',
      lastUpdated: '2024-10-22',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-haiku-4-5': {
      displayName: 'Claude Haiku 4.5',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-1': {
      displayName: 'Claude Opus 4.1',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-5': {
      displayName: 'Claude Opus 4.5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-6': {
      displayName: 'Claude Opus 4.6',
      description: 'High-end Claude for difficult coding, planning, and slower expert reasoning',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05-31',
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-7': {
      displayName: 'Claude Opus 4.7',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-4-8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-opus-5': {
      displayName: 'Claude Opus 5',
      description: 'Strongest Claude Opus model for coding, agents, and professional work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-05',
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4': {
      displayName: 'Claude Sonnet 4',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4-5': {
      displayName: 'Claude Sonnet 4.5',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-07-31',
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-4-6': {
      displayName: 'Claude Sonnet 4.6',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-02-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash (New)',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-flash-free': {
      displayName: 'DeepSeek V4 Flash Free (New)',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gemini-3-flash': {
      displayName: 'Gemini 3 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3-pro': {
      displayName: 'Gemini 3 Pro',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-11-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.1-pro': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.5-flash-lite': {
      displayName: 'Gemini 3.5 Flash Lite',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'glm-4.6': {
      displayName: 'GLM-4.6',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.7-free': {
      displayName: 'GLM-4.7 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5-free': {
      displayName: 'GLM-5 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-5': {
      displayName: 'GPT-5',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5-codex': {
      displayName: 'GPT-5 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-09-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5-nano': {
      displayName: 'GPT-5 Nano',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.1': {
      displayName: 'GPT-5.1',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.1-codex': {
      displayName: 'GPT-5.1 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.1-codex-max': {
      displayName: 'GPT-5.1 Codex Max',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.1-codex-mini': {
      displayName: 'GPT-5.1 Codex Mini',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2': {
      displayName: 'GPT-5.2',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'gpt-5.2-codex': {
      displayName: 'GPT-5.2 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-01-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.3-codex': {
      displayName: 'GPT-5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.3-codex-spark': {
      displayName: 'GPT-5.3 Codex Spark',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 128000,
      inputLimit: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-5.4': {
      displayName: 'GPT-5.4',
      description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4-mini': {
      displayName: 'GPT-5.4 Mini',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4-nano': {
      displayName: 'GPT-5.4 Nano',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.4-pro': {
      displayName: 'GPT-5.4 Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: false,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.5': {
      displayName: 'GPT-5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.5-pro': {
      displayName: 'GPT-5.5 Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: false,
      lastUpdated: '2026-04-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna',
      description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-sol': {
      displayName: 'GPT-5.6 Sol',
      description:
        'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'gpt-5.6-terra': {
      displayName: 'GPT-5.6 Terra',
      description: 'Balanced GPT-5.6 model for capable, cost-efficient everyday work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'grok-build-0.1': {
      displayName: 'Grok Build 0.1',
      description: 'Grok coding model for agentic engineering, edits, and codebase workflows',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-05-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'grok-code': {
      displayName: 'Grok Code Fast 1',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2025-08-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hy3-free': {
      displayName: 'Hy3 Free',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 190000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-06-26',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hy3-preview-free': {
      displayName: 'Hy3 preview Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2': {
      displayName: 'Kimi K2',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2-thinking': {
      displayName: 'Kimi K2 Thinking',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.5-free': {
      displayName: 'Kimi K2.5 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'laguna-s-2.1-free': {
      displayName: 'Laguna S 2.1 Free',
      description: 'Agentic coding model from Poolside in the XS size class for local deployment',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      structuredOutput: false,
      lastUpdated: '2026-07-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ling-2.6-flash-free': {
      displayName: 'Ling 2.6 Flash Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262100,
      maxOutputTokens: 32800,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-04-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ling-3.0-flash-free': {
      displayName: 'Ling-3.0-flash Free',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'longcat-2.0-free': {
      displayName: 'LongCat-2.0 Free',
      description:
        'Meituan LongCat-2.0, a reasoning model with tool calling and a 1M-token context window',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      lastUpdated: '2026-06-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-flash-free': {
      displayName: 'MiMo V2 Flash Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-omni-free': {
      displayName: 'MiMo V2 Omni Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'mimo-v2-pro-free': {
      displayName: 'MiMo V2 Pro Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5-free': {
      displayName: 'MiMo V2.5 Free',
      description: 'MiMo omni model for text, image, video, audio, and agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'minimax-m2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.1-free': {
      displayName: 'MiniMax-M2.1 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.5-free': {
      displayName: 'MiniMax-M2.5 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 512000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'minimax-m3-free': {
      displayName: 'MiniMax-M3 Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-31',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nemotron-3-super-free': {
      displayName: 'Nemotron 3 Super Free',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02',
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nemotron-3-ultra-free': {
      displayName: 'Nemotron 3 Ultra Free',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02',
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'north-mini-code-free': {
      displayName: 'North Mini Code Free',
      description: 'Cohere coding model for practical software engineering and agentic edits',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-09-23',
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3-coder': {
      displayName: 'Qwen3 Coder',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus-free': {
      displayName: 'Qwen3.6 Plus Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'ring-2.6-1t-free': {
      displayName: 'Ring 2.6 1T Free',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262000,
      maxOutputTokens: 66000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-05-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'trinity-large-preview-free': {
      displayName: 'Trinity Large Preview',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-01-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'opencode-go': {
    'deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash (New)',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Flagship DeepSeek model for coding, reasoning, and agentic work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 202752,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 202752,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna (2x usage)',
      description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    hy3: {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mimo-v2-omni': {
      displayName: 'MiMo V2 Omni',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'mimo-v2-pro': {
      displayName: 'MiMo V2 Pro',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5': {
      displayName: 'MiMo V2.5',
      description: 'MiMo omni model for text, image, video, audio, and agents',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'mimo-v2.5-pro': {
      displayName: 'MiMo V2.5 Pro',
      description: 'MiMo pro model for strong multimodal reasoning and agent execution',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1048576,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax-m3': {
      displayName: 'MiniMax-M3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-31',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen3.8-max': {
      displayName: 'Qwen3.8 Max',
      description:
        '2.4-trillion-parameter multimodal flagship for coding, professional work, and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://opencode.ai/docs/zen',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  openrouter: {
    '~anthropic/claude-fable-latest': {
      displayName: 'Claude Fable Latest',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    '~anthropic/claude-haiku-latest': {
      displayName: 'Anthropic Claude Haiku Latest',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    '~anthropic/claude-opus-latest': {
      displayName: 'Claude Opus Latest',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    '~anthropic/claude-sonnet-latest': {
      displayName: 'Anthropic Claude Sonnet Latest',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    '~deepseek/deepseek-v4-flash-latest': {
      displayName: 'DeepSeek V4 Flash Latest',
      description: 'Fast DeepSeek model for efficient chat, coding help, and agent loops',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-08-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    '~google/gemini-flash-latest': {
      displayName: 'Google Gemini Flash Latest',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01-01',
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    '~google/gemini-pro-latest': {
      displayName: 'Google Gemini Pro Latest',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['audio', 'pdf', 'image', 'text'], output: ['text'] },
    },
    '~moonshotai/kimi-latest': {
      displayName: 'MoonshotAI Kimi Latest',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    '~openai/gpt-latest': {
      displayName: 'OpenAI GPT Latest',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    '~openai/gpt-mini-latest': {
      displayName: 'OpenAI GPT Mini Latest',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    '~x-ai/grok-latest': {
      displayName: 'Grok Latest',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 500000,
      maxOutputTokens: 1000000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'ai21/jamba-large-1.7': {
      displayName: 'Jamba Large 1.7',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: false,
      lastUpdated: '2025-08-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'aion-labs/aion-2.0': {
      displayName: 'Aion-2.0',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-02-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'aion-labs/aion-3.0': {
      displayName: 'Aion-3.0',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'aion-labs/aion-3.0-mini': {
      displayName: 'Aion-3.0-Mini',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'aion-labs/aion-rp-llama-3.1-8b': {
      displayName: 'Aion-RP 1.0 (8B)',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: false,
      lastUpdated: '2025-02-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'allenai/olmo-3-32b-think': {
      displayName: 'Olmo 3 32B Think',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2025-11-21',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'amazon/nova-2-lite-v1': {
      displayName: 'Nova 2 Lite',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65535,
      structuredOutput: false,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'amazon/nova-lite-v1': {
      displayName: 'Nova Lite 1.0',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 300000,
      maxOutputTokens: 5120,
      knowledgeCutoff: '2024-10-31',
      structuredOutput: false,
      lastUpdated: '2024-12-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'amazon/nova-micro-v1': {
      displayName: 'Nova Micro 1.0',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 5120,
      knowledgeCutoff: '2024-10-31',
      structuredOutput: false,
      lastUpdated: '2024-12-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'amazon/nova-premier-v1': {
      displayName: 'Nova Premier 1.0',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 32000,
      structuredOutput: false,
      lastUpdated: '2025-10-31',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'amazon/nova-pro-v1': {
      displayName: 'Nova Pro 1.0',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 300000,
      maxOutputTokens: 5120,
      knowledgeCutoff: '2024-10-31',
      structuredOutput: false,
      lastUpdated: '2024-12-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'anthracite-org/magnum-v4-72b': {
      displayName: 'Magnum v4 72B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 16384,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2024-10-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'anthropic/claude-3-haiku': {
      displayName: 'Claude 3 Haiku',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-08-31',
      structuredOutput: false,
      lastUpdated: '2024-03-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'anthropic/claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-haiku-4.5': {
      displayName: 'Claude Haiku 4.5 (latest)',
      description: 'Fast Claude lane for lightweight agents, office tasks, and responsive chat',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      structuredOutput: true,
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4': {
      displayName: 'Claude Opus 4',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: false,
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.1': {
      displayName: 'Claude Opus 4.1 (latest)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: false,
      lastUpdated: '2025-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.5': {
      displayName: 'Claude Opus 4.5 (latest)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.6': {
      displayName: 'Claude Opus 4.6',
      description: 'High-end Claude for difficult coding, planning, and slower expert reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05-31',
      structuredOutput: true,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.7': {
      displayName: 'Claude Opus 4.7',
      description: 'Stronger Opus tier for advanced software work and high-stakes reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.7-fast': {
      displayName: 'Claude Opus 4.7 (Fast)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.8-fast': {
      displayName: 'Claude Opus 4.8 (Fast)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-5': {
      displayName: 'Claude Opus 5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-05',
      structuredOutput: true,
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-5-fast': {
      displayName: 'Claude Opus 5 (Fast)',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-05',
      structuredOutput: true,
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4': {
      displayName: 'Claude Sonnet 4',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: false,
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.5': {
      displayName: 'Claude Sonnet 4.5 (latest)',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-07-31',
      structuredOutput: true,
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.6': {
      displayName: 'Claude Sonnet 4.6',
      description:
        'Claude workhorse for coding agents, careful analysis, and production cost control',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      structuredOutput: true,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'arcee-ai/trinity-large-thinking': {
      displayName: 'Trinity Large Thinking',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'arcee-ai/virtuoso-large': {
      displayName: 'Virtuoso Large',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: false,
      lastUpdated: '2025-05-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'baidu/ernie-4.5-vl-424b-a47b': {
      displayName: 'ERNIE 4.5 VL 424B A47B ',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 123000,
      maxOutputTokens: 16000,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: false,
      lastUpdated: '2025-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'bytedance-seed/seed-1.6': {
      displayName: 'Seed 1.6',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-12-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'bytedance-seed/seed-1.6-flash': {
      displayName: 'Seed 1.6 Flash',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-12-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'bytedance-seed/seed-2.0-lite': {
      displayName: 'Seed-2.0-Lite',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'bytedance-seed/seed-2.0-mini': {
      displayName: 'Seed-2.0-Mini',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'bytedance/ui-tars-1.5-7b': {
      displayName: 'UI-TARS 7B ',
      description: 'Multimodal model for analyzing text, images, documents, and rich media',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: true,
      lastUpdated: '2025-07-22',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'cognitivecomputations/dolphin-mistral-24b-venice-edition': {
      displayName: 'Uncensored',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-04-30',
      structuredOutput: false,
      lastUpdated: '2025-07-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/command-a': {
      displayName: 'Command A',
      description: 'Cohere command model for multilingual enterprise agents, tools, and chat',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-03-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/command-r-08-2024': {
      displayName: 'Command R',
      description: 'Cohere retrieval model for long-context chat and enterprise RAG workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      structuredOutput: true,
      lastUpdated: '2024-08-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/command-r-plus-08-2024': {
      displayName: 'Command R+',
      description: "Cohere's RAG workhorse for long-context enterprise search and tool use",
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      structuredOutput: true,
      lastUpdated: '2024-08-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/command-r7b-12-2024': {
      displayName: 'Command R7B',
      description: 'Cohere retrieval model for long-context chat and enterprise RAG workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2024-06-01',
      structuredOutput: true,
      lastUpdated: '2024-12-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/north-mini-code:free': {
      displayName: 'North Mini Code (free)',
      description: 'Cohere coding model for practical software engineering and agentic edits',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      structuredOutput: false,
      lastUpdated: '2026-06-17',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepcogito/cogito-v2.1-671b': {
      displayName: 'Cogito v2.1 671B',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-chat': {
      displayName: 'DeepSeek Chat',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 16000,
      knowledgeCutoff: '2025-09',
      structuredOutput: true,
      lastUpdated: '2026-02-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-chat-v3-0324': {
      displayName: 'DeepSeek V3 0324',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-07-31',
      structuredOutput: true,
      lastUpdated: '2025-03-24',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-chat-v3.1': {
      displayName: 'DeepSeek V3.1',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-r1': {
      displayName: 'DeepSeek-R1',
      description:
        'Classic open reasoning model for transparent math, coding, and deliberate problem solving',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 16000,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-r1-0528': {
      displayName: 'R1 0528',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-05-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-r1-distill-llama-70b': {
      displayName: 'R1 Distill Llama 70B',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-07-31',
      structuredOutput: false,
      lastUpdated: '2025-01-23',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.1-terminus': {
      displayName: 'DeepSeek V3.1 Terminus',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-09-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2': {
      displayName: 'DeepSeek V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2025-12-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2-exp': {
      displayName: 'DeepSeek V3.2 Exp',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 163840,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-07-31',
      structuredOutput: true,
      lastUpdated: '2025-09-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemini-2.5-flash': {
      displayName: 'Gemini 2.5 Flash',
      description: 'Fast Gemini workhorse for multimodal apps where latency and price matter',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65535,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-2.5-flash-image': {
      displayName: 'Nano Banana',
      description:
        'Nano Banana image model for fast generation, edits, and character-consistent assets',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      structuredOutput: true,
      lastUpdated: '2025-08-26',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-2.5-flash-lite': {
      displayName: 'Gemini 2.5 Flash-Lite',
      description: 'Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65535,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-2.5-pro': {
      displayName: 'Gemini 2.5 Pro',
      description: "Google's proven reasoning model for coding, math, and multimodal analysis",
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-2.5-pro-preview': {
      displayName: 'Gemini 2.5 Pro Preview 06-05',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: true,
      lastUpdated: '2025-06-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['pdf', 'image', 'text', 'audio'], output: ['text'] },
    },
    'google/gemini-2.5-pro-preview-05-06': {
      displayName: 'Gemini 2.5 Pro Preview 05-06',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65535,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: true,
      lastUpdated: '2025-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    'google/gemini-3-flash-preview': {
      displayName: 'Gemini 3 Flash Preview',
      description:
        'New Gemini flash lane bringing frontier-style multimodal reasoning to cheaper runs',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3-pro-image': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3-pro-image-preview': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-11-20',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-image': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-image-preview': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['image', 'text'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-lite': {
      displayName: 'Gemini 3.1 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.1-flash-lite-image': {
      displayName: 'Nano Banana 2 Lite',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-lite-preview': {
      displayName: 'Gemini 3.1 Flash Lite Preview',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Reasoning-first Gemini preview for agentic coding and complex problem solving',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.1-pro-preview-customtools': {
      displayName: 'Gemini 3.1 Pro Preview Custom Tools',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.5-flash-lite': {
      displayName: 'Gemini 3.5 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemma-2-27b-it': {
      displayName: 'Gemma 2 27B',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8192,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2024-07-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-3-12b-it': {
      displayName: 'Gemma 3 12B',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-03-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3-27b-it': {
      displayName: 'Gemma 3 27B',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-03-12',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3-4b-it': {
      displayName: 'Gemma 3 4B',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-03-13',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'google/gemma-3n-e4b-it': {
      displayName: 'Gemma 3n 4B',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-26b-a4b-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'google/gemma-4-26b-a4b-it:free': {
      displayName: 'Gemma 4 26B A4B  (free)',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'google/gemma-4-31b-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'google/gemma-4-31b-it:free': {
      displayName: 'Gemma 4 31B (free)',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'google/lyria-3-clip-preview': {
      displayName: 'Lyria 3 Clip Preview',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'google/lyria-3-pro-preview': {
      displayName: 'Lyria 3 Pro Preview',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-03-25',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'audio'] },
    },
    'gryphe/mythomax-l2-13b': {
      displayName: 'MythoMax 13B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8192,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-06-30',
      structuredOutput: true,
      lastUpdated: '2023-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ibm-granite/granite-4.0-h-micro': {
      displayName: 'Granite 4.0 Micro',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: false,
      lastUpdated: '2025-10-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ibm-granite/granite-4.1-8b': {
      displayName: 'Granite 4.1 8B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inception/mercury-2': {
      displayName: 'Mercury 2',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 50000,
      structuredOutput: true,
      lastUpdated: '2026-03-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-2.6-1t': {
      displayName: 'Ling-2.6-1T',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-2.6-flash': {
      displayName: 'Ling-2.6-flash',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-3.0-flash': {
      displayName: 'Ling-3.0-flash',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-3.0-tiny:free': {
      displayName: 'Ling 3.0 Tiny (free)',
      description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-08-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ring-2.6-1t': {
      displayName: 'Ring-2.6-1T',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-05-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kwaipilot/kat-coder-air-v2.5': {
      displayName: 'KAT-Coder-Air V2.5',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 80000,
      structuredOutput: true,
      lastUpdated: '2026-07-10',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kwaipilot/kat-coder-pro-v2': {
      displayName: 'KAT-Coder-Pro V2',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 80000,
      structuredOutput: true,
      lastUpdated: '2026-03-27',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kwaipilot/kat-coder-pro-v2.5': {
      displayName: 'KAT-Coder-Pro V2.5',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 80000,
      structuredOutput: true,
      lastUpdated: '2026-07-10',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mancer/weaver': {
      displayName: 'Weaver (alpha)',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8000,
      maxOutputTokens: 2000,
      knowledgeCutoff: '2023-06-30',
      structuredOutput: true,
      lastUpdated: '2023-08-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meituan/longcat-2.0': {
      displayName: 'LongCat 2.0',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048756,
      maxOutputTokens: 262144,
      structuredOutput: false,
      lastUpdated: '2026-07-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-3.1-70b-instruct': {
      displayName: 'Llama 3.1 70B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-3.1-8b-instruct': {
      displayName: 'Llama 3.1 8B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-3.2-1b-instruct': {
      displayName: 'Llama 3.2 1B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 60000,
      maxOutputTokens: 60000,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: false,
      lastUpdated: '2024-09-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-3.2-3b-instruct': {
      displayName: 'Llama 3.2 3B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-09-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-3.3-70b-instruct': {
      displayName: 'Llama-3.3-70B-Instruct',
      description: 'Popular open Llama workhorse for multilingual chat, coding, and self-hosting',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12',
      structuredOutput: true,
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/llama-4-maverick': {
      displayName: 'Llama 4 Maverick',
      description: 'Open multimodal Llama model for strong reasoning and fast responses',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta-llama/llama-4-scout': {
      displayName: 'Llama 4 Scout',
      description: 'Open multimodal Llama model for long-context analysis and efficient agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1310720,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: true,
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta-llama/llama-guard-4-12b': {
      displayName: 'Llama Guard 4 12B',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: false,
      lastUpdated: '2025-04-30',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'meta/muse-spark-1.1': {
      displayName: 'Muse Spark 1.1',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    'meta/muse-spark-1.2': {
      displayName: 'Muse Spark 1.2',
      description:
        'Muse Spark 1.2 is a coding-focused update to Muse Spark 1.1 with improvements in code generation, complex debugging, codebase understanding, and end-to-end developer workflows.',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    'microsoft/phi-4': {
      displayName: 'Phi 4',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 16384,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2025-01-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'microsoft/wizardlm-2-8x22b': {
      displayName: 'WizardLM-2 8x22B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65535,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-04-30',
      structuredOutput: false,
      lastUpdated: '2024-04-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-01': {
      displayName: 'MiniMax-01',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000192,
      maxOutputTokens: 1000192,
      knowledgeCutoff: '2024-03-31',
      structuredOutput: false,
      lastUpdated: '2025-01-15',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'minimax/minimax-m1': {
      displayName: 'MiniMax M1',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 40000,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: false,
      lastUpdated: '2025-06-17',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2': {
      displayName: 'MiniMax-M2',
      description: 'Efficient open MiniMax model built for coding agents and tool-heavy workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2-her': {
      displayName: 'MiniMax M2-her',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 2048,
      structuredOutput: false,
      lastUpdated: '2026-01-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.1': {
      displayName: 'MiniMax-M2.1',
      description: 'Earlier MiniMax agent model for practical coding and productivity tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Prior MiniMax coding model for agent workflows, office edits, and automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 196608,
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.7': {
      displayName: 'MiniMax-M2.7',
      description:
        'Open MiniMax flagship for coding agents, office automation, and complex environments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 512000,
      structuredOutput: true,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/codestral-2508': {
      displayName: 'Codestral 2508',
      description: 'Mistral coding model for code completion, generation, and developer workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-08-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'mistralai/ministral-14b-2512': {
      displayName: 'Ministral 3 14B 2512',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/ministral-3b-2512': {
      displayName: 'Ministral 3 3B 2512',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/ministral-8b-2512': {
      displayName: 'Ministral 3 8B 2512',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-large': {
      displayName: 'Mistral Large',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-11-30',
      structuredOutput: true,
      lastUpdated: '2024-02-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-large-2407': {
      displayName: 'Mistral Large 2407',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-03-31',
      structuredOutput: true,
      lastUpdated: '2024-11-19',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-large-2512': {
      displayName: 'Mistral Large 3',
      description:
        "Mistral's largest general model for enterprise agents, coding, and multilingual reasoning",
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-11',
      structuredOutput: true,
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-medium-3': {
      displayName: 'Mistral Medium 3',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-05-07',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-medium-3-5': {
      displayName: 'Mistral Medium 3.5',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-medium-3.1': {
      displayName: 'Mistral Medium 3.1',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: true,
      lastUpdated: '2025-08-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-nemo': {
      displayName: 'Mistral Nemo',
      description: 'Efficient Mistral-NVIDIA open model for multilingual chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2024-07-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/mistral-saba': {
      displayName: 'Saba',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-02-17',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'mistralai/mistral-small-24b-instruct-2501': {
      displayName: 'Mistral Small 3',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-10-31',
      structuredOutput: true,
      lastUpdated: '2025-01-30',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistralai/mistral-small-2603': {
      displayName: 'Mistral Small 4',
      description: 'Fast Mistral production model for chat, extraction, and cost-sensitive agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-06',
      structuredOutput: true,
      lastUpdated: '2026-03-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-small-3.1-24b-instruct': {
      displayName: 'Mistral Small 3.1 24B',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2023-10-31',
      structuredOutput: false,
      lastUpdated: '2025-03-17',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistralai/mistral-small-3.2-24b-instruct': {
      displayName: 'Mistral Small 3.2 24B',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-10-31',
      structuredOutput: true,
      lastUpdated: '2025-06-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'mistralai/mixtral-8x22b-instruct': {
      displayName: 'Mixtral 8x22B Instruct',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-01-31',
      structuredOutput: true,
      lastUpdated: '2024-04-17',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'mistralai/voxtral-small-24b-2507': {
      displayName: 'Voxtral Small 24B 2507',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      structuredOutput: true,
      lastUpdated: '2025-10-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'audio', 'pdf'], output: ['text'] },
    },
    'moonshotai/kimi-k2': {
      displayName: 'Kimi K2 0711',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 100352,
      knowledgeCutoff: '2024-12-31',
      structuredOutput: false,
      lastUpdated: '2025-07-11',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-0905': {
      displayName: 'Kimi K2 0905',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 100352,
      knowledgeCutoff: '2024-12-31',
      structuredOutput: true,
      lastUpdated: '2025-09-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-thinking': {
      displayName: 'Kimi K2 Thinking',
      description:
        'Thinking Kimi model for slower research passes, planning, and hard technical questions',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 100352,
      knowledgeCutoff: '2024-08',
      structuredOutput: true,
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description:
        'Earlier Kimi frontier model for long-context agents, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k3': {
      displayName: 'Kimi K3',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'morph/morph-v3-fast': {
      displayName: 'Morph V3 Fast',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 81920,
      maxOutputTokens: 38000,
      structuredOutput: false,
      lastUpdated: '2025-07-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'morph/morph-v3-large': {
      displayName: 'Morph V3 Large',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-07-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nex-agi/nex-n2-mini': {
      displayName: 'Nex-N2-Mini',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-06-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nex-agi/nex-n2-pro': {
      displayName: 'Nex-N2-Pro',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: false,
      lastUpdated: '2026-06-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nousresearch/hermes-3-llama-3.1-405b': {
      displayName: 'Hermes 3 405B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-08-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nousresearch/hermes-3-llama-3.1-70b': {
      displayName: 'Hermes 3 70B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-08-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nousresearch/hermes-4-405b': {
      displayName: 'Hermes 4 405B',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: false,
      lastUpdated: '2025-08-26',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nousresearch/hermes-4-70b': {
      displayName: 'Hermes 4 70B',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-08-31',
      structuredOutput: false,
      lastUpdated: '2025-08-26',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-30b-a3b': {
      displayName: 'Nemotron 3 Nano 30B A3B',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2025-12-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-30b-a3b:free': {
      displayName: 'Nemotron 3 Nano 30B A3B (free)',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: false,
      lastUpdated: '2025-12-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free': {
      displayName: 'Nemotron 3 Nano Omni (free)',
      description: 'Open Nemotron omni model combining reasoning with text, vision, and audio',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-04-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'nvidia/nemotron-3-super-120b-a12b': {
      displayName: 'Nemotron 3 Super 120B A12B',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-super-120b-a12b:free': {
      displayName: 'Nemotron 3 Super (free)',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-ultra-550b-a55b': {
      displayName: 'Nemotron 3 Ultra 550B A55B',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 512288,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-ultra-550b-a55b:free': {
      displayName: 'Nemotron 3 Ultra (free)',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3.5-content-safety:free': {
      displayName: 'Nemotron 3.5 Content Safety (free)',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      structuredOutput: false,
      lastUpdated: '2026-06-04',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/nemotron-nano-12b-v2-vl:free': {
      displayName: 'Nemotron Nano 12B 2 VL (free)',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-10-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/nemotron-nano-9b-v2:free': {
      displayName: 'Nemotron Nano 9B V2 (free)',
      description: 'Compact Nemotron model for efficient reasoning and deployable AI agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2025-08-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-3.5-turbo': {
      displayName: 'GPT-3.5-turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 16385,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09-01',
      structuredOutput: true,
      lastUpdated: '2023-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-3.5-turbo-0613': {
      displayName: 'GPT-3.5 Turbo (older v0613)',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 4095,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09-30',
      structuredOutput: true,
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-3.5-turbo-16k': {
      displayName: 'GPT-3.5 Turbo 16k',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 16385,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09-30',
      structuredOutput: true,
      lastUpdated: '2023-08-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-3.5-turbo-instruct': {
      displayName: 'GPT-3.5 Turbo Instruct',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 4095,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09-30',
      structuredOutput: true,
      lastUpdated: '2023-09-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-4': {
      displayName: 'GPT-4',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8191,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-11',
      structuredOutput: true,
      lastUpdated: '2024-04-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-4-turbo': {
      displayName: 'GPT-4 Turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: true,
      lastUpdated: '2024-04-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-4-turbo-preview': {
      displayName: 'GPT-4 Turbo Preview',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-4.1': {
      displayName: 'GPT-4.1',
      description:
        'Long-lived GPT workhorse for coding, instruction following, and production apps',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4.1-mini': {
      displayName: 'GPT-4.1 mini',
      description: 'Affordable GPT-4.1 lane for fast coding help and structured extraction',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4.1-nano': {
      displayName: 'GPT-4.1 nano',
      description: 'Tiny GPT-4.1 option for classification, routing, and very high-volume tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o': {
      displayName: 'GPT-4o',
      description: 'Omni-era GPT for multimodal chat, practical coding, and general assistants',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-08-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-2024-05-13': {
      displayName: 'GPT-4o (2024-05-13)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-05-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-2024-08-06': {
      displayName: 'GPT-4o (2024-08-06)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-08-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-2024-11-20': {
      displayName: 'GPT-4o (2024-11-20)',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-11-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-mini': {
      displayName: 'GPT-4o mini',
      description: 'Small omni GPT for cheap multimodal assistance and production-scale traffic',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-07-18',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-mini-2024-07-18': {
      displayName: 'GPT-4o-mini (2024-07-18)',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-10-31',
      structuredOutput: true,
      lastUpdated: '2024-07-18',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5': {
      displayName: 'GPT-5',
      description: 'Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5-image': {
      displayName: 'GPT-5 Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10-01',
      structuredOutput: true,
      lastUpdated: '2025-10-14',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['image', 'text', 'pdf'], output: ['image', 'text'] },
    },
    'openai/gpt-5-image-mini': {
      displayName: 'GPT-5 Image Mini',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2025-10-16',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['pdf', 'image', 'text'], output: ['image', 'text'] },
    },
    'openai/gpt-5-mini': {
      displayName: 'GPT-5 Mini',
      description: 'Small GPT-5 for responsive agents, coding help, and everyday automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5-nano': {
      displayName: 'GPT-5 Nano',
      description: 'Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5-pro': {
      displayName: 'GPT-5 Pro',
      description: 'Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-10-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1': {
      displayName: 'GPT-5.1',
      description: 'Sharper GPT-5 generation for coding, product work, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-codex': {
      displayName: 'GPT-5.1 Codex',
      description: 'Codex GPT for repository edits, code review, and practical software agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.1-codex-max': {
      displayName: 'GPT-5.1 Codex Max',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.1-codex-mini': {
      displayName: 'GPT-5.1 Codex mini',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.2': {
      displayName: 'GPT-5.2',
      description:
        'Reliable GPT generation for broad coding, writing, and tool-assisted product work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.2-chat': {
      displayName: 'GPT-5.2 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-10',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.2-codex': {
      displayName: 'GPT-5.2 Codex',
      description:
        'Code-specialist GPT for repository edits, reviews, and long-running software agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.2-pro': {
      displayName: 'GPT-5.2 Pro',
      description: 'Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.3-chat': {
      displayName: 'GPT-5.3 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.3-codex': {
      displayName: 'GPT-5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4': {
      displayName: 'GPT-5.4',
      description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4-image-2': {
      displayName: 'GPT-5.4 Image 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 272000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['image', 'text'] },
    },
    'openai/gpt-5.4-mini': {
      displayName: 'GPT-5.4 mini',
      description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.4-nano': {
      displayName: 'GPT-5.4 nano',
      description: 'Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.4-pro': {
      displayName: 'GPT-5.4 Pro',
      description: 'More exact GPT-5.4 tier for demanding professional reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5': {
      displayName: 'GPT-5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5-pro': {
      displayName: 'GPT-5.5 Pro',
      description: 'Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-luna-pro': {
      displayName: 'GPT-5.6 Luna Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-sol': {
      displayName: 'GPT-5.6 Sol',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-sol-pro': {
      displayName: 'GPT-5.6 Sol Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-terra': {
      displayName: 'GPT-5.6 Terra',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-terra-pro': {
      displayName: 'GPT-5.6 Terra Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-audio': {
      displayName: 'GPT Audio',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-audio-mini': {
      displayName: 'GPT Audio Mini',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-chat-latest': {
      displayName: 'GPT Chat Latest',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-05-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b:free': {
      displayName: 'gpt-oss-20b (free)',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-safeguard-20b': {
      displayName: 'gpt-oss-safeguard-20b',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2025-10-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/o1': {
      displayName: 'o1',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-12-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o1-pro': {
      displayName: 'o1-pro',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2025-03-19',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o3': {
      displayName: 'o3',
      description: 'Deliberate o-series reasoner for hard math, coding, and multi-step analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o3-mini': {
      displayName: 'o3-mini',
      description: 'Smaller o-series reasoner for economical coding, math, and planning tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-01-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'openai/o3-mini-high': {
      displayName: 'o3 Mini High',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-10-31',
      structuredOutput: true,
      lastUpdated: '2025-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['text', 'pdf'], output: ['text'] },
    },
    'openai/o3-pro': {
      displayName: 'o3-pro',
      description: 'High-effort o3 tier for difficult technical reasoning and careful answers',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-06-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'pdf', 'image'], output: ['text'] },
    },
    'openai/o4-mini': {
      displayName: 'o4-mini',
      description: 'Fast o-series model for compact reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/o4-mini-high': {
      displayName: 'o4 Mini High',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openrouter/auto': {
      displayName: 'Auto Router',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      structuredOutput: true,
      lastUpdated: '2023-11-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text', 'image'] },
    },
    'openrouter/bodybuilder': {
      displayName: 'Body Builder (beta)',
      description:
        'Preview model for early access evaluation, prototyping, and compatibility testing',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-12-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openrouter/free': {
      displayName: 'Free Models Router',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      inputLimit: 200000,
      maxOutputTokens: 8000,
      structuredOutput: true,
      lastUpdated: '2026-02-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openrouter/fusion': {
      displayName: 'Fusion',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openrouter/pareto-code': {
      displayName: 'Pareto Code Router',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 2000000,
      maxOutputTokens: 200000,
      structuredOutput: false,
      lastUpdated: '2026-04-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'perceptron/perceptron-mk1': {
      displayName: 'Perceptron Mk1',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 8192,
      structuredOutput: true,
      lastUpdated: '2026-05-12',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar': {
      displayName: 'Sonar',
      description: 'Sonar search model for current answers, retrieval, and citation-backed chat',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 127072,
      maxOutputTokens: 127072,
      structuredOutput: false,
      lastUpdated: '2025-01-27',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar-deep-research': {
      displayName: 'Sonar Deep Research',
      description: 'Sonar search model for current answers, retrieval, and citation-backed chat',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-03-07',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'perplexity/sonar-pro': {
      displayName: 'Sonar Pro',
      description: 'Advanced Sonar search model for deeper research and cited synthesis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 8000,
      structuredOutput: false,
      lastUpdated: '2025-03-07',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar-pro-search': {
      displayName: 'Sonar Pro Search',
      description: 'Advanced Sonar search model for deeper research and cited synthesis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 200000,
      maxOutputTokens: 8000,
      structuredOutput: true,
      lastUpdated: '2025-10-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar-reasoning-pro': {
      displayName: 'Sonar Reasoning Pro',
      description: 'Web-grounded reasoning model for multi-step research and cited answers',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-03-07',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'poolside/laguna-s-2.1': {
      displayName: 'Laguna S 2.1',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2026-07-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'poolside/laguna-s-2.1:free': {
      displayName: 'Laguna S 2.1 (free)',
      description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'poolside/laguna-xs-2.1': {
      displayName: 'Laguna XS 2.1',
      description: 'Agentic coding model from Poolside in the XS size class for local deployment',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'poolside/laguna-xs-2.1:free': {
      displayName: 'Laguna XS 2.1 (free)',
      description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-2.5-72b-instruct': {
      displayName: 'Qwen2.5 72B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2024-09-19',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-2.5-7b-instruct': {
      displayName: 'Qwen2.5 7B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2024-10-16',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-2.5-coder-32b-instruct': {
      displayName: 'Qwen2.5 Coder 32B Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: false,
      lastUpdated: '2024-11-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-plus': {
      displayName: 'Qwen Plus',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-09-11',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-plus-2025-07-28': {
      displayName: 'Qwen Plus 0728',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-09-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen-plus-2025-07-28:thinking': {
      displayName: 'Qwen Plus 0728 (thinking)',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-09-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen2.5-vl-72b-instruct': {
      displayName: 'Qwen2.5 VL 72B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2025-02-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-14b': {
      displayName: 'Qwen3 14B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-04-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-235b-a22b': {
      displayName: 'Qwen3 235B-A22B',
      description: 'Large open Qwen MoE for multilingual reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-235b-a22b-2507': {
      displayName: 'Qwen3 235B A22B Instruct 2507',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: true,
      lastUpdated: '2025-07-21',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-235b-a22b-thinking-2507': {
      displayName: 'Qwen3 235B A22B Thinking 2507',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: false,
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-30b-a3b': {
      displayName: 'Qwen3 30B A3B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: false,
      lastUpdated: '2025-04-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-30b-a3b-instruct-2507': {
      displayName: 'Qwen3 30B A3B Instruct 2507',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: true,
      lastUpdated: '2025-07-29',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-30b-a3b-thinking-2507': {
      displayName: 'Qwen3 30B A3B Thinking 2507',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 81920,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: false,
      lastUpdated: '2025-08-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-32b': {
      displayName: 'Qwen3 32B',
      description: 'Dense open Qwen model for self-hosted chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-8b': {
      displayName: 'Qwen3 8B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: false,
      lastUpdated: '2025-04-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder': {
      displayName: 'Qwen3 Coder 480B A35B',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-06-30',
      structuredOutput: true,
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder-30b-a3b-instruct': {
      displayName: 'Qwen3-Coder 30B-A3B Instruct',
      description: 'Smaller Qwen coder for efficient local agents and repo-level fixes',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder-flash': {
      displayName: 'Qwen3 Coder Flash',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder-next': {
      displayName: 'Qwen3 Coder Next',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-09',
      structuredOutput: true,
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-coder-plus': {
      displayName: 'Qwen3 Coder Plus',
      description: 'Hosted Qwen coder for software agents, repo edits, and long-context code',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-max': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen3 model for coding agents, complex reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-max-thinking': {
      displayName: 'Qwen3 Max Thinking',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-next-80b-a3b-instruct': {
      displayName: 'Qwen3-Next 80B-A3B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-next-80b-a3b-thinking': {
      displayName: 'Qwen3-Next 80B-A3B (Thinking)',
      description: 'Efficient Qwen thinking model for local reasoning, math, and coding agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-vl-235b-a22b-instruct': {
      displayName: 'Qwen3 VL 235B A22B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-09-23',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-vl-235b-a22b-thinking': {
      displayName: 'Qwen3 VL 235B A22B Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-09-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-vl-30b-a3b-instruct': {
      displayName: 'Qwen3 VL 30B A3B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-10-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-vl-30b-a3b-thinking': {
      displayName: 'Qwen3 VL 30B A3B Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-10-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-vl-32b-instruct': {
      displayName: 'Qwen3 VL 32B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-10-23',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3-vl-8b-instruct': {
      displayName: 'Qwen3 VL 8B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-10-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'qwen/qwen3-vl-8b-thinking': {
      displayName: 'Qwen3 VL 8B Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-10-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'qwen/qwen3.5-122b-a10b': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 81920,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-27b': {
      displayName: 'Qwen3.5 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-35b-a3b': {
      displayName: 'Qwen3.5 35B-A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-397b-a17b': {
      displayName: 'Qwen3.5 397B-A17B',
      description: 'Large open Qwen multimodal MoE for visual agents and long technical tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-9b': {
      displayName: 'Qwen3.5 9B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-flash-02-23': {
      displayName: 'Qwen3.5-Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-02-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-plus-02-15': {
      displayName: 'Qwen3.5 Plus 2026-02-15',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-plus-20260420': {
      displayName: 'Qwen3.5 Plus 2026-04-20',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.6-27b': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.6-35b-a3b': {
      displayName: 'Qwen3.6 35B-A3B',
      description: 'Open multimodal Qwen MoE for local agents that need vision, audio, and code',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.6-flash': {
      displayName: 'Qwen3.6 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-04-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.6-max-preview': {
      displayName: 'Qwen3.6 Max Preview',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.6-plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Earlier Qwen multimodal workhorse for million-token agent and document tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.7-flash': {
      displayName: 'Qwen3.7 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      inputLimit: 991000,
      maxOutputTokens: 65536,
      structuredOutput: false,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.8-max': {
      displayName: 'Qwen3.8 Max',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-08-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'rekaai/reka-edge': {
      displayName: 'Reka Edge',
      description: 'Multimodal model for analyzing text, images, documents, and rich media',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 16384,
      maxOutputTokens: 16384,
      structuredOutput: true,
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'rekaai/reka-flash-3': {
      displayName: 'Reka Flash 3',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01-31',
      structuredOutput: true,
      lastUpdated: '2025-03-12',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'relace/relace-apply-3': {
      displayName: 'Relace Apply 3',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-09-26',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'relace/relace-search': {
      displayName: 'Relace Search',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 128000,
      structuredOutput: false,
      lastUpdated: '2025-12-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'sakana/fugu-ultra': {
      displayName: 'Fugu Ultra',
      description: 'Quality-first multi-agent model for hard research, analysis, and competitions',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'sao10k/l3-lunaris-8b': {
      displayName: 'Llama 3 8B Lunaris',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 8192,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-08-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'sao10k/l3.1-euryale-70b': {
      displayName: 'Llama 3.1 Euryale 70B v2.2',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-08-28',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'sao10k/l3.3-euryale-70b': {
      displayName: 'Llama 3.3 Euryale 70B',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-12-31',
      structuredOutput: true,
      lastUpdated: '2024-12-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun/step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun/step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      structuredOutput: true,
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'tencent/hunyuan-a13b-instruct': {
      displayName: 'Hunyuan A13B Instruct',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-03-31',
      structuredOutput: true,
      lastUpdated: '2025-07-08',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'tencent/hy3': {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'tencent/hy3-preview': {
      displayName: 'Hy3 preview',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: false,
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thedrummer/cydonia-24b-v4.1': {
      displayName: 'Cydonia 24B V4.1',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-04-30',
      structuredOutput: true,
      lastUpdated: '2025-09-27',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thedrummer/rocinante-12b': {
      displayName: 'Rocinante 12B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-04-30',
      structuredOutput: true,
      lastUpdated: '2024-09-30',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thedrummer/skyfall-36b-v2': {
      displayName: 'Skyfall 36B V2',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06-30',
      structuredOutput: true,
      lastUpdated: '2025-03-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thedrummer/unslopnemo-12b': {
      displayName: 'UnslopNemo 12B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1024000,
      maxOutputTokens: 1024000,
      knowledgeCutoff: '2024-04-30',
      structuredOutput: true,
      lastUpdated: '2024-11-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thinkingmachines/inkling': {
      displayName: 'Inkling',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: false,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'thinkingmachines/inkling-small': {
      displayName: 'Inkling Small',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 524288,
      maxOutputTokens: 262144,
      structuredOutput: false,
      lastUpdated: '2026-07-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'undi95/remm-slerp-l2-13b': {
      displayName: 'ReMM SLERP 13B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 6144,
      maxOutputTokens: 2048,
      knowledgeCutoff: '2023-06-30',
      structuredOutput: true,
      lastUpdated: '2023-07-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'upstage/solar-pro-3': {
      displayName: 'Solar Pro 3',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-01-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'writer/palmyra-x5': {
      displayName: 'Palmyra X5',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1040000,
      maxOutputTokens: 8192,
      structuredOutput: false,
      lastUpdated: '2026-01-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'x-ai/grok-4.20': {
      displayName: 'Grok 4.20',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      knowledgeCutoff: '2025-09-01',
      structuredOutput: true,
      lastUpdated: '2026-03-31',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-4.20-multi-agent': {
      displayName: 'Grok 4.20 Multi-Agent',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      knowledgeCutoff: '2025-09-01',
      structuredOutput: true,
      lastUpdated: '2026-03-31',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-4.3': {
      displayName: 'Grok 4.3',
      description:
        "xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-build-0.1': {
      displayName: 'Grok Build 0.1',
      description: 'Fast Grok coding model tuned for agentic engineering and iterative edits',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xiaomi/mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'xiaomi/mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1050000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.5': {
      displayName: 'GLM-4.5',
      description: 'Hybrid-reasoning GLM release that made the 4.5 line broadly useful',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.5-air': {
      displayName: 'GLM-4.5-Air',
      description: 'Lighter GLM-4.5 variant for fast coding assistance and cheaper agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.5v': {
      displayName: 'GLM-4.5V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 65536,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-08-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'z-ai/glm-4.6': {
      displayName: 'GLM-4.6',
      description: 'Late GLM-4 workhorse for coding agents, reasoning, and structured tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.6v': {
      displayName: 'GLM-4.6V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      structuredOutput: false,
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'z-ai/glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Mature GLM model for dependable coding, reasoning, and structured agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.7-flash': {
      displayName: 'GLM-4.7-Flash',
      description: 'Budget GLM lane for fast coding help, routing, and everyday automation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      structuredOutput: true,
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5-turbo': {
      displayName: 'GLM-5-Turbo',
      description: 'Faster GLM-5 lane for coding agents that need lower latency',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.1': {
      displayName: 'GLM-5.1',
      description:
        'Strong GLM coding model for agentic engineering, terminals, and repository generation',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5v-turbo': {
      displayName: 'GLM-5V-Turbo',
      description: 'Fast GLM vision model for screenshots, documents, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://openrouter.ai/models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
  },
  siliconflow: {
    'baidu/ERNIE-4.5-300B-A47B': {
      displayName: 'baidu/ERNIE-4.5-300B-A47B',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'ByteDance-Seed/Seed-OSS-36B-Instruct': {
      displayName: 'ByteDance-Seed/Seed-OSS-36B-Instruct',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-R1': {
      displayName: 'deepseek-ai/DeepSeek-R1',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3': {
      displayName: 'deepseek-ai/DeepSeek-V3',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.1': {
      displayName: 'deepseek-ai/DeepSeek-V3.1',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.1-Terminus': {
      displayName: 'deepseek-ai/DeepSeek-V3.1-Terminus',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.2': {
      displayName: 'deepseek-ai/DeepSeek-V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-12-03',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3.2-Exp': {
      displayName: 'deepseek-ai/DeepSeek-V3.2-Exp',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 164000,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-26B-A4B-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-31B-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionAI/Ling-flash-2.0': {
      displayName: 'inclusionAI/Ling-flash-2.0',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.5': {
      displayName: 'MiniMaxAI/MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 197000,
      maxOutputTokens: 131000,
      structuredOutput: false,
      lastUpdated: '2026-06-15',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.5': {
      displayName: 'moonshotai/Kimi-K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.6': {
      displayName: 'moonshotai/Kimi-K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2026-06-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'openai/gpt-oss-120b',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 8000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'openai/gpt-oss-20b',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 8000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen2.5-72B-Instruct': {
      displayName: 'Qwen/Qwen2.5-72B-Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 33000,
      maxOutputTokens: 4000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen2.5-7B-Instruct': {
      displayName: 'Qwen/Qwen2.5-7B-Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 33000,
      maxOutputTokens: 4000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-14B': {
      displayName: 'Qwen/Qwen3-14B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B-Thinking-2507': {
      displayName: 'Qwen/Qwen3-235B-A22B-Thinking-2507',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-30B-A3B-Instruct-2507': {
      displayName: 'Qwen/Qwen3-30B-A3B-Instruct-2507',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-32B': {
      displayName: 'Qwen/Qwen3-32B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-8B': {
      displayName: 'Qwen/Qwen3-8B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-30B-A3B-Instruct': {
      displayName: 'Qwen/Qwen3-Coder-30B-A3B-Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-480B-A35B-Instruct': {
      displayName: 'Qwen/Qwen3-Coder-480B-A35B-Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-235B-A22B-Instruct': {
      displayName: 'Qwen/Qwen3-VL-235B-A22B-Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-235B-A22B-Thinking': {
      displayName: 'Qwen/Qwen3-VL-235B-A22B-Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-30B-A3B-Instruct': {
      displayName: 'Qwen/Qwen3-VL-30B-A3B-Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-30B-A3B-Thinking': {
      displayName: 'Qwen/Qwen3-VL-30B-A3B-Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-32B-Instruct': {
      displayName: 'Qwen/Qwen3-VL-32B-Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-32B-Thinking': {
      displayName: 'Qwen/Qwen3-VL-32B-Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3-VL-8B-Instruct': {
      displayName: 'Qwen/Qwen3-VL-8B-Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-122B-A10B': {
      displayName: 'Qwen3.5 122B-A10B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-27B': {
      displayName: 'Qwen3.5 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-35B-A3B': {
      displayName: 'Qwen3.5 35B-A3B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-397B-A17B': {
      displayName: 'Qwen3.5 397B-A17B',
      description: 'Large open Qwen multimodal MoE for visual agents and long technical tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-02-15',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-9B': {
      displayName: 'Qwen/Qwen3.5-9B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.6-27B': {
      displayName: 'Qwen3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.6-35B-A3B': {
      displayName: 'Qwen3.6 35B-A3B',
      description: 'Open multimodal Qwen MoE for local agents that need vision, audio, and code',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun-ai/Step-3.5-Flash': {
      displayName: 'stepfun-ai/Step-3.5-Flash',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'tencent/Hunyuan-A13B-Instruct': {
      displayName: 'tencent/Hunyuan-A13B-Instruct',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'tencent/Hy3-preview': {
      displayName: 'Hy3 preview',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-4.5-Air': {
      displayName: 'zai-org/GLM-4.5-Air',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 131000,
      maxOutputTokens: 131000,
      structuredOutput: true,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5': {
      displayName: 'zai-org/GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 205000,
      maxOutputTokens: 205000,
      structuredOutput: true,
      lastUpdated: '2026-06-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.1': {
      displayName: 'zai-org/GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 205000,
      maxOutputTokens: 205000,
      structuredOutput: true,
      lastUpdated: '2026-04-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 1049000,
      maxOutputTokens: 262000,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5V-Turbo': {
      displayName: 'zai-org/GLM-5V-Turbo',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://cloud.siliconflow.com/models',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  stepfun: {
    'step-1-32k': {
      displayName: 'Step 1 (32K)',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 32768,
      inputLimit: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-2-16k': {
      displayName: 'Step 2 (16K)',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 16384,
      inputLimit: 16384,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-06-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash-2603': {
      displayName: 'Step 3.5 Flash 2603',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-06-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'step-tts-2': {
      displayName: 'Step TTS 2',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'stepaudio-2.5-asr': {
      displayName: 'StepAudio 2.5 ASR',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'stepaudio-2.5-tts': {
      displayName: 'StepAudio 2.5 TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
  },
  'stepfun-ai': {
    'step-1-32k': {
      displayName: 'Step 1 (32K)',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 32768,
      inputLimit: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-2-16k': {
      displayName: 'Step 2 (16K)',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 16384,
      inputLimit: 16384,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-06',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-06-15',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash-2603': {
      displayName: 'Step 3.5 Flash 2603',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-06-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'step-tts-2': {
      displayName: 'Step TTS 2',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'stepaudio-2.5-asr': {
      displayName: 'StepAudio 2.5 ASR',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'stepaudio-2.5-tts': {
      displayName: 'StepAudio 2.5 TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/overview/concept',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
  },
  'stepfun-ai-step-plan': {
    'step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash-2603': {
      displayName: 'Step 3.5 Flash 2603',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
  },
  'stepfun-step-plan': {
    'step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.5-flash-2603': {
      displayName: 'Step 3.5 Flash 2603',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'step-router-v1': {
      displayName: 'Step Router v1',
      description: 'StepFun routing model that dispatches requests to the appropriate Step model.',
      lifecycle: 'active',
      docsUrl: 'https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  togetherai: {
    'deepcogito/cogito-v2-1-671b': {
      displayName: 'Cogito v2.1 671B',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 163840,
      maxOutputTokens: 163840,
      lastUpdated: '2025-11-13',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-R1': {
      displayName: 'DeepSeek-R1',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 163839,
      maxOutputTokens: 163839,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-03-24',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3': {
      displayName: 'DeepSeek-V3',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V3-1': {
      displayName: 'DeepSeek V3.1',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-08',
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek-ai/DeepSeek-V4-Pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Flagship DeepSeek model for coding, reasoning, and agentic work',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 512000,
      maxOutputTokens: 384000,
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'essentialai/Rnj-1-Instruct': {
      displayName: 'Rnj-1 Instruct',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-3n-E4B-it': {
      displayName: 'Gemma 3N E4B Instruct',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemma-4-31B-it': {
      displayName: 'Gemma 4 31B Instruct',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'LiquidAI/LFM2-24B-A2B': {
      displayName: 'LFM2-24B-A2B',
      description:
        'Open-weight instruction model for adaptable chat and self-hosted production workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      lastUpdated: '2026-02-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/Llama-3.3-70B-Instruct-Turbo': {
      displayName: 'Llama 3.3 70B',
      description: 'Compact Llama instruction model for fast chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta-llama/Meta-Llama-3-8B-Instruct-Lite': {
      displayName: 'Meta Llama 3 8B Instruct Lite',
      description: 'Compact Llama instruction model for fast chat and local deployment',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2024-04-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M2.7': {
      displayName: 'MiniMax-M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'MiniMaxAI/MiniMax-M3': {
      displayName: 'MiniMax-M3',
      description: 'MiniMax multimodal coding model for long-context reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 524288,
      maxOutputTokens: 250000,
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.5': {
      displayName: 'Kimi K2.5',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 131000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/Kimi-K2.7-Code': {
      displayName: 'Kimi K2.7 Code',
      description: 'Kimi coding model for software agents, refactors, and repository reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-14',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/Kimi-K3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/nemotron-3-ultra-550b-a55b': {
      displayName: 'Nemotron 3 Ultra 550B A55B',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 512300,
      maxOutputTokens: 512300,
      structuredOutput: true,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-08',
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'pearl-ai/gemma-4-31b-it': {
      displayName: 'Pearl AI Gemma 4 31B Instruct',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen2.5-7B-Instruct-Turbo': {
      displayName: 'Qwen 2.5 7B Instruct Turbo',
      description: 'Efficient Qwen model for fast chat, extraction, and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      structuredOutput: true,
      lastUpdated: '2024-09-19',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-235B-A22B-Instruct-2507-tput': {
      displayName: 'Qwen3 235B A22B Instruct 2507 FP8',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-07',
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8': {
      displayName: 'Qwen3 Coder 480B A35B Instruct',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3-Coder-Next-FP8': {
      displayName: 'Qwen3 Coder Next FP8',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2026-02-03',
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.5-397B-A17B': {
      displayName: 'Qwen3.5 397B A17B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 130000,
      lastUpdated: '2026-06-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.5-9B': {
      displayName: 'Qwen3.5 9B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      structuredOutput: true,
      lastUpdated: '2026-03-03',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'Qwen/Qwen3.6-Plus': {
      displayName: 'Qwen3.6 Plus',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 1000000,
      maxOutputTokens: 500000,
      lastUpdated: '2026-04-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'Qwen/Qwen3.7-Max': {
      displayName: 'Qwen3.7 Max',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 1000000,
      maxOutputTokens: 500000,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thinkingmachines/Inkling': {
      displayName: 'Inkling',
      description:
        'Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 524288,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['max', 'xhigh', 'high', 'medium', 'low', 'none'] },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'zai-org/GLM-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'deprecated',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 202752,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-11',
      structuredOutput: true,
      lastUpdated: '2026-07-02',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai-org/GLM-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.together.ai/docs/serverless-models',
      contextWindow: 262144,
      maxOutputTokens: 164000,
      structuredOutput: true,
      lastUpdated: '2026-06-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'tencent-coding-plan': {
    'glm-5': {
      displayName: 'GLM-5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 202752,
      maxOutputTokens: 16384,
      lastUpdated: '2026-02-11',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hunyuan-2.0-instruct': {
      displayName: 'Tencent HY 2.0 Instruct',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hunyuan-2.0-thinking': {
      displayName: 'Tencent HY 2.0 Think',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hunyuan-t1': {
      displayName: 'Hunyuan-T1',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-08',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hunyuan-turbos': {
      displayName: 'Hunyuan-TurboS',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kimi-k2.5': {
      displayName: 'Kimi-K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'minimax-m2.5': {
      displayName: 'MiniMax-M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 204800,
      maxOutputTokens: 32768,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'tc-code-latest': {
      displayName: 'Auto',
      description: 'Automatic model router for matching prompts to suitable backends and budgets',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1772/128947',
      contextWindow: 131072,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-08',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'tencent-token-plan': {
    hy3: {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1823/130060',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'tencent-tokenhub': {
    hy3: {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1823/130050',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'hy3-preview': {
      displayName: 'Hy3 preview',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://cloud.tencent.com/document/product/1823/130050',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  vercel: {
    'alibaba/qwen-3-14b': {
      displayName: 'Qwen3-14B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 40960,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen-3-235b': {
      displayName: 'Qwen3 235B A22B Instruct 2507',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen-3-30b': {
      displayName: 'Qwen3-30B-A3B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 40960,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen-3-32b': {
      displayName: 'Qwen 3.32B',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen-3.6-max-preview': {
      displayName: 'Qwen 3.6 Max Preview',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 240000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-235b-a22b-thinking': {
      displayName: 'Qwen3 235B A22B Thinking 2507',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3-coder': {
      displayName: 'Qwen3 Coder 480B A35B Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-coder-30b-a3b': {
      displayName: 'Qwen 3 Coder 30B A3B Instruct',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-coder-next': {
      displayName: 'Qwen3 Coder Next',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2025-09',
      structuredOutput: true,
      lastUpdated: '2026-02-03',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-coder-plus': {
      displayName: 'Qwen3 Coder Plus',
      description: 'Hosted Qwen coder for software agents, repo edits, and long-context code',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-embedding-0.6b': {
      displayName: 'Qwen3 Embedding 0.6B',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      lastUpdated: '2025-11-14',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-embedding-4b': {
      displayName: 'Qwen3 Embedding 4B',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      lastUpdated: '2025-06-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-embedding-8b': {
      displayName: 'Qwen3 Embedding 8B',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32768,
      maxOutputTokens: 32768,
      lastUpdated: '2025-06-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-max': {
      displayName: 'Qwen3 Max',
      description: 'Flagship Qwen3 model for coding agents, complex reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-max-preview': {
      displayName: 'Qwen3 Max Preview',
      description: 'Flagship Qwen model for complex reasoning, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-max-thinking': {
      displayName: 'Qwen 3 Max Thinking',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-next-80b-a3b-instruct': {
      displayName: 'Qwen3 Next 80B A3B Instruct',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-next-80b-a3b-thinking': {
      displayName: 'Qwen3 Next 80B A3B Thinking',
      description: 'Efficient Qwen thinking model for local reasoning, math, and coding agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2025-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3-vl-235b-a22b-instruct': {
      displayName: 'Qwen3 VL 235B A22B Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 129024,
      lastUpdated: '2026-05-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'alibaba/qwen3-vl-instruct': {
      displayName: 'Qwen3 VL Instruct',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 129024,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-24',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'alibaba/qwen3-vl-thinking': {
      displayName: 'Qwen3 VL Thinking',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2025-09-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.5-flash': {
      displayName: 'Qwen 3.5 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-02-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.5-plus': {
      displayName: 'Qwen 3.5 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-02-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.6-27b': {
      displayName: 'Qwen 3.6 27B',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.6-plus': {
      displayName: 'Qwen 3.6 Plus',
      description: 'Earlier Qwen multimodal workhorse for million-token agent and document tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.7-flash': {
      displayName: 'Qwen 3.7 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 991000,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.7-max': {
      displayName: 'Qwen 3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 991000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'alibaba/qwen3.7-plus': {
      displayName: 'Qwen 3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'alibaba/qwen3.8-max': {
      displayName: 'Qwen 3.8 Max',
      description:
        'Preview Qwen flagship for million-token multimodal reasoning and long-horizon agentic workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-07-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'alibaba/wan-v2.5-t2v-preview': {
      displayName: 'Wan v2.5 Text-to-Video Preview',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-09-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.6-i2v': {
      displayName: 'Wan v2.6 Image-to-Video',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.6-i2v-flash': {
      displayName: 'Wan v2.6 Image-to-Video Flash',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.6-r2v': {
      displayName: 'Wan v2.6 Reference-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.6-r2v-flash': {
      displayName: 'Wan v2.6 Reference-to-Video Flash',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.6-t2v': {
      displayName: 'Wan v2.6 Text-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.7-r2v': {
      displayName: 'Wan v2.7 Reference-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'alibaba/wan-v2.7-t2v': {
      displayName: 'Wan v2.7 Text-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'amazon/nova-2-lite': {
      displayName: 'Nova 2 Lite',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-12-01',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'amazon/nova-lite': {
      displayName: 'Nova Lite',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 300000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-12-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'amazon/nova-micro': {
      displayName: 'Nova Micro',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-12-03',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'amazon/nova-pro': {
      displayName: 'Nova Pro',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 300000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-12-03',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'amazon/titan-embed-text-v2': {
      displayName: 'Titan Text Embeddings V2',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'anthropic/claude-3-haiku': {
      displayName: 'Claude Haiku 3',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-08-31',
      lastUpdated: '2024-03-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'anthropic/claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-haiku-4.5': {
      displayName: 'Claude Haiku 4.5',
      description: 'Fast Claude lane for lightweight agents, office tasks, and responsive chat',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-02-28',
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4': {
      displayName: 'Claude Opus 4',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.5': {
      displayName: 'Claude Opus 4.5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.6': {
      displayName: 'Claude Opus 4.6',
      description: 'High-end Claude for difficult coding, planning, and slower expert reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05-31',
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.7': {
      displayName: 'Claude Opus 4.7',
      description: 'Stronger Opus tier for advanced software work and high-stakes reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.8-fast': {
      displayName: 'Claude Opus 4.8 (Fast)',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-5': {
      displayName: 'Claude Opus 5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-05',
      lastUpdated: '2026-07-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4': {
      displayName: 'Claude Sonnet 4',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2025-03-31',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.5': {
      displayName: 'Claude Sonnet 4.5',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-07-31',
      structuredOutput: true,
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.6': {
      displayName: 'Claude Sonnet 4.6',
      description:
        'Claude workhorse for coding agents, careful analysis, and production cost control',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'], toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'arcee-ai/trinity-large-thinking': {
      displayName: 'Trinity Large Thinking',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262100,
      maxOutputTokens: 80000,
      lastUpdated: '2026-04-03',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'arcee-ai/trinity-mini': {
      displayName: 'Trinity Mini',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'bfl/flux-2-flex': {
      displayName: 'FLUX.2 [flex]',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-2-klein-4b': {
      displayName: 'FLUX.2 [klein] 4B',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-2-klein-9b': {
      displayName: 'FLUX.2 [klein] 9B',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-2-max': {
      displayName: 'FLUX.2 [max]',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 67300,
      maxOutputTokens: 67300,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-2-pro': {
      displayName: 'FLUX.2 [pro]',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 67300,
      maxOutputTokens: 67300,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-3-video': {
      displayName: 'Flux 3',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-08-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'bfl/flux-kontext-max': {
      displayName: 'FLUX.1 Kontext Max',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2025-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-kontext-pro': {
      displayName: 'FLUX.1 Kontext Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2025-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-pro-1.0-fill': {
      displayName: 'FLUX.1 Fill [pro]',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2024-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-pro-1.1': {
      displayName: 'FLUX1.1 [pro]',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2024-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bfl/flux-pro-1.1-ultra': {
      displayName: 'FLUX1.1 [pro] Ultra',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2024-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bytedance/seed-1.6': {
      displayName: 'Seed 1.6',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'bytedance/seed-1.8': {
      displayName: 'Seed 1.8',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'bytedance/seedance-2.0': {
      displayName: 'Seedance 2.0',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'bytedance/seedance-2.0-fast': {
      displayName: 'Seedance 2.0 Fast',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'bytedance/seedance-v1.0-pro': {
      displayName: 'Seedance v1.0 Pro',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-06-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'bytedance/seedance-v1.0-pro-fast': {
      displayName: 'Seedance v1.0 Pro Fast',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-10-31',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'bytedance/seedance-v1.5-pro': {
      displayName: 'Seedance v1.5 Pro',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'bytedance/seedream-4.0': {
      displayName: 'Seedream 4.0',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-08-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bytedance/seedream-4.5': {
      displayName: 'Seedream 4.5',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-11-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bytedance/seedream-5.0-lite': {
      displayName: 'Seedream 5.0 Lite',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-01-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'bytedance/seedream-5.0-pro': {
      displayName: 'Seedream 5.0 Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'cohere/command-a': {
      displayName: 'Command A',
      description: 'Cohere command model for multilingual enterprise agents, tools, and chat',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-03-13',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/embed-v4.0': {
      displayName: 'Embed v4.0',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 1536,
      lastUpdated: '2025-04-15',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/rerank-v3.5': {
      displayName: 'Cohere Rerank 3.5',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 4096,
      maxOutputTokens: 4096,
      lastUpdated: '2024-12-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/rerank-v4-fast': {
      displayName: 'Cohere Rerank 4 Fast',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2025-12-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'cohere/rerank-v4-pro': {
      displayName: 'Cohere Rerank 4 Pro',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2025-12-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-r1': {
      displayName: 'DeepSeek-R1',
      description: 'DeepSeek reasoning model for multi-step analysis, math, coding, and tools',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-05-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3': {
      displayName: 'DeepSeek V3 0324',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 163840,
      maxOutputTokens: 163840,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2024-12-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.1': {
      displayName: 'DeepSeek-V3.1',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 163840,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-08-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.1-terminus': {
      displayName: 'DeepSeek V3.1 Terminus',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-07',
      lastUpdated: '2025-09-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2': {
      displayName: 'DeepSeek V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-07',
      structuredOutput: true,
      lastUpdated: '2025-12-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2-thinking': {
      displayName: 'DeepSeek V3.2 Thinking',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2024-07',
      lastUpdated: '2025-12-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-flash-0731': {
      displayName: 'DeepSeek V4 Flash 0731',
      description:
        'Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-07-31',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048600,
      maxOutputTokens: 1048600,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'fish-audio/s1': {
      displayName: 'S1',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-10-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'fish-audio/s2-pro': {
      displayName: 'S2 Pro',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'fish-audio/s2.1-pro': {
      displayName: 'S2.1 Pro',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'fish-audio/transcribe-1': {
      displayName: 'Transcribe-1',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'google/gemini-2.5-flash': {
      displayName: 'Gemini 2.5 Flash',
      description: 'Fast Gemini workhorse for multimodal apps where latency and price matter',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-2.5-flash-image': {
      displayName: 'Nano Banana (Gemini 2.5 Flash Image)',
      description:
        'Nano Banana image model for fast generation, edits, and character-consistent assets',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32768,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2025-08-26',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-2.5-flash-lite': {
      displayName: 'Gemini 2.5 Flash Lite',
      description: 'Lean Gemini 2.5 lane for cheap multimodal traffic and quick agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-2.5-pro': {
      displayName: 'Gemini 2.5 Pro',
      description: "Google's proven reasoning model for coding, math, and multimodal analysis",
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3-flash': {
      displayName: 'Gemini 3 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 65000,
      knowledgeCutoff: '2025-03',
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-3-pro-image': {
      displayName: 'Nano Banana Pro',
      description: 'Nano Banana Pro for higher-fidelity image generation and design-heavy edits',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 65536,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-image': {
      displayName: 'Nano Banana 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-image-preview': {
      displayName: 'Gemini 3.1 Flash Image Preview (Nano Banana 2)',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-26',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-flash-lite': {
      displayName: 'Gemini 3.1 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 65000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.1-flash-lite-image': {
      displayName: 'Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 65536,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2025-01',
      structuredOutput: false,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text', 'image'] },
    },
    'google/gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Reasoning-first Gemini preview for agentic coding and complex problem solving',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.5-flash-lite': {
      displayName: 'Gemini 3.5 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 65000,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.6-flash': {
      displayName: 'Gemini 3.6 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-03',
      structuredOutput: true,
      lastUpdated: '2026-07-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemini-embedding-001': {
      displayName: 'Gemini Embedding 001',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      knowledgeCutoff: '2025-05',
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemini-embedding-2': {
      displayName: 'Gemini Embedding 2',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      knowledgeCutoff: '2025-11',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemini-omni-flash-preview': {
      displayName: 'Gemini Omni Flash Preview',
      description: 'Omni-modal model for text, vision, audio, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 57920,
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemma-4-26b-a4b-it': {
      displayName: 'Gemma 4 26B A4B IT',
      description: 'Open Gemma instruction model for efficient chat and self-hosted deployments',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/gemma-4-31b-it': {
      displayName: 'Gemma 4 31B IT',
      description: 'Largest Gemma 4 instruction model for open, self-hosted chat and reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-02',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'google/imagen-4.0-fast-generate-001': {
      displayName: 'Imagen 4 Fast',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 480,
      maxOutputTokens: 0,
      lastUpdated: '2025-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'google/imagen-4.0-generate-001': {
      displayName: 'Imagen 4',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 480,
      maxOutputTokens: 0,
      lastUpdated: '2025-05-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'google/imagen-4.0-ultra-generate-001': {
      displayName: 'Imagen 4 Ultra',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 480,
      maxOutputTokens: 0,
      lastUpdated: '2025-05-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'google/text-embedding-005': {
      displayName: 'Text Embedding 005',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/text-multilingual-embedding-002': {
      displayName: 'Text Multilingual Embedding 002',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/veo-3.0-fast-generate-001': {
      displayName: 'Veo 3.0 Fast Generate',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'google/veo-3.0-generate-001': {
      displayName: 'Veo 3.0',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'google/veo-3.1-fast-generate-001': {
      displayName: 'Veo 3.1 Fast Generate',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'google/veo-3.1-generate-001': {
      displayName: 'Veo 3.1',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'google/veo-3.1-lite-generate-001': {
      displayName: 'Veo 3.1 Lite Generate',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-02',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'inception/mercury-2': {
      displayName: 'Mercury 2',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-03-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inception/mercury-coder-small': {
      displayName: 'Mercury Coder Small Beta',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 16384,
      lastUpdated: '2025-02-26',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-3.0-flash': {
      displayName: 'Ling 3.0 Flash',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      lastUpdated: '2026-08-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ling-3.0-tiny-free': {
      displayName: 'Ling 3.0 Tiny (Free)',
      description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      lastUpdated: '2026-08-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'interfaze/interfaze-beta': {
      displayName: 'Interfaze Beta',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 32000,
      lastUpdated: '2026-04-29',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'klingai/kling-v2.5-turbo-i2v': {
      displayName: 'Kling v2.5 Turbo Image-to-Video',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v2.5-turbo-t2v': {
      displayName: 'Kling v2.5 Turbo Text-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-09-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v2.6-i2v': {
      displayName: 'Kling v2.6 Image-to-Video',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v2.6-motion-control': {
      displayName: 'Kling v2.6 Motion Control',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v2.6-t2v': {
      displayName: 'Kling v2.6 Text-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-12-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v3.0-i2v': {
      displayName: 'Kling v3.0 Image-to-Video',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v3.0-motion-control': {
      displayName: 'Kling v3.0 Motion Control',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-04',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'klingai/kling-v3.0-t2v': {
      displayName: 'Kling v3.0 Text-to-Video',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-05',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'kwaipilot/kat-coder-air-v2.5': {
      displayName: 'Kat Coder Air V2.5',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 80000,
      lastUpdated: '2026-07-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'kwaipilot/kat-coder-pro-v1': {
      displayName: 'KAT-Coder-Pro V1',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-10-24',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kwaipilot/kat-coder-pro-v2': {
      displayName: 'Kat Coder Pro V2',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2026-03-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kwaipilot/kat-coder-pro-v2.5': {
      displayName: 'Kat Coder Pro V2.5',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 80000,
      lastUpdated: '2026-07-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/llama-3.1-70b': {
      displayName: 'Llama 3.1 70B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.1-8b': {
      displayName: 'Llama 3.1 8B Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-3.3-70b': {
      displayName: 'Llama-3.3-70B-Instruct',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      lastUpdated: '2024-12-06',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'meta/llama-4-maverick': {
      displayName: 'Llama-4-Maverick-17B-128E-Instruct-FP8',
      description: 'Open multimodal Llama model for strong reasoning and fast responses',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/llama-4-scout': {
      displayName: 'Llama-4-Scout-17B-16E-Instruct-FP8',
      description: 'Open multimodal Llama model for long-context analysis and efficient agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-04-05',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'meta/muse-spark-1.1': {
      displayName: 'Muse Spark 1.1',
      description: 'Open Llama instruction model for multilingual chat, reasoning, and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'meta/muse-spark-1.2': {
      displayName: 'Muse Spark 1.2',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      structuredOutput: true,
      lastUpdated: '2026-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'meta/muse-spark-1.2-contributor': {
      displayName: 'Muse Spark 1.2 Contributor',
      description: 'Open Llama multimodal model for image understanding and text reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1048576,
      maxOutputTokens: 1048576,
      lastUpdated: '2026-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'minimax/minimax-h3': {
      displayName: 'MiniMax H3',
      description: 'Video model for prompt-guided generation, editing, and motion workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-30',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: [] },
    },
    'minimax/minimax-m2': {
      displayName: 'MiniMax M2',
      description: 'Efficient open MiniMax model built for coding agents and tool-heavy workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 205000,
      maxOutputTokens: 205000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.1': {
      displayName: 'MiniMax M2.1',
      description: 'Earlier MiniMax agent model for practical coding and productivity tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.1-lightning': {
      displayName: 'MiniMax M2.1 Lightning',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.5': {
      displayName: 'MiniMax M2.5',
      description: 'Prior MiniMax coding model for agent workflows, office edits, and automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131000,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.5-highspeed': {
      displayName: 'MiniMax M2.5 High Speed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131000,
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.7': {
      displayName: 'Minimax M2.7',
      description:
        'Open MiniMax flagship for coding agents, office automation, and complex environments',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131000,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.7-highspeed': {
      displayName: 'MiniMax M2.7 High Speed',
      description: 'Low-latency M2.7 variant for interactive coding plans and agent loops',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 204800,
      maxOutputTokens: 131100,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m3': {
      displayName: 'MiniMax M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistral/codestral': {
      displayName: 'Codestral (latest)',
      description: 'Mistral code model for completions, refactors, and developer IDE workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-01-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/codestral-embed': {
      displayName: 'Codestral Embed',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2025-05-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/devstral-2': {
      displayName: 'Devstral 2',
      description:
        'Mistral coding agent model for repository tasks and software engineering workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/devstral-small-2': {
      displayName: 'Devstral Small 2',
      description:
        'Mistral coding agent model for repository tasks and software engineering workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-05-07',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/magistral-medium': {
      displayName: 'Magistral Medium (latest)',
      description: 'Mistral reasoning model for transparent analysis, math, and complex decisions',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2025-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/magistral-small': {
      displayName: 'Magistral Small',
      description: 'Mistral reasoning model for transparent analysis, math, and complex decisions',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2025-03-17',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/ministral-14b': {
      displayName: 'Ministral 14B',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-01',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'mistral/ministral-3b': {
      displayName: 'Ministral 3B (latest)',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-10-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/ministral-8b': {
      displayName: 'Ministral 8B (latest)',
      description:
        'Compact Mistral model for edge, latency-sensitive, and cost-efficient workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-10-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/mistral-embed': {
      displayName: 'Mistral Embed',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2023-12-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mistral/mistral-large-3': {
      displayName: 'Mistral Large 3',
      description: 'Flagship Mistral model for advanced reasoning, coding, and multilingual work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-02',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/mistral-medium': {
      displayName: 'Mistral Medium 3.1',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-05-07',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/mistral-medium-3.5': {
      displayName: 'Mistral Medium Latest',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2026-05-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/mistral-nemo': {
      displayName: 'Mistral Nemo',
      description: 'Mistral model for multilingual chat, reasoning, and tool-assisted workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-04',
      lastUpdated: '2024-07-01',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/mistral-small': {
      displayName: 'Mistral Small (latest)',
      description: 'Efficient Mistral model for fast chat, extraction, and production assistants',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 4000,
      knowledgeCutoff: '2025-06',
      lastUpdated: '2026-03-16',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'mistral/pixtral-12b': {
      displayName: 'Pixtral 12B',
      description: 'Mistral vision-language model for image understanding and multimodal chat',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09',
      lastUpdated: '2024-09-01',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2': {
      displayName: 'Kimi K2 Instruct',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      lastUpdated: '2025-09-05',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-thinking': {
      displayName: 'Kimi K2 Thinking',
      description:
        'Thinking Kimi model for slower research passes, planning, and hard technical questions',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 216144,
      maxOutputTokens: 216144,
      knowledgeCutoff: '2024-08',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description:
        'Earlier Kimi frontier model for long-context agents, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262114,
      maxOutputTokens: 262114,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Multimodal Kimi workhorse for agent loops, coding tasks, and visual context',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262000,
      maxOutputTokens: 262000,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-04-21',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'moonshotai/kimi-k2.7-code-highspeed': {
      displayName: 'Kimi K2.7 Code High Speed',
      description: 'Lower-latency Kimi Code variant for interactive edits and coding-agent loops',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'moonshotai/kimi-k3': {
      displayName: 'Kimi K3',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'moonshotai/kimi-k3-fast': {
      displayName: 'Kimi K3 Fast',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'high', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'morph/morph-v3-fast': {
      displayName: 'Morph v3 Fast',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 16000,
      maxOutputTokens: 16000,
      lastUpdated: '2024-08-15',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'morph/morph-v3-large': {
      displayName: 'Morph v3 Large',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2024-08-15',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-nano-30b-a3b': {
      displayName: 'Nemotron 3 Nano 30B A3B',
      description: 'Small Nemotron 3 MoE for efficient coding, math, and long-context agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-15',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-super-120b-a12b': {
      displayName: 'NVIDIA Nemotron 3 Super 120B A12B',
      description:
        'Nemotron middle tier for collaborative agents and high-volume reasoning workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      lastUpdated: '2026-03-11',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-3-ultra-550b-a55b': {
      displayName: 'Nemotron 3 Ultra',
      description: 'Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 65000,
      lastUpdated: '2026-06-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'nvidia/nemotron-nano-12b-v2-vl': {
      displayName: 'Nvidia Nemotron Nano 12B V2 VL',
      description: 'Nemotron multimodal model for visual reasoning and agentic AI workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-10-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'nvidia/nemotron-nano-9b-v2': {
      displayName: 'Nvidia Nemotron Nano 9B V2',
      description: 'Compact Nemotron model for efficient reasoning and deployable AI agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-08-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-3.5-turbo': {
      displayName: 'GPT-3.5 Turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 16385,
      inputLimit: 12289,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2021-09',
      structuredOutput: false,
      lastUpdated: '2023-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-4-turbo': {
      displayName: 'GPT-4 Turbo',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'deprecated',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 4096,
      knowledgeCutoff: '2023-12',
      structuredOutput: false,
      lastUpdated: '2024-04-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-4.1': {
      displayName: 'GPT-4.1',
      description:
        'Long-lived GPT workhorse for coding, instruction following, and production apps',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4.1-mini': {
      displayName: 'GPT-4.1 mini',
      description: 'Affordable GPT-4.1 lane for fast coding help and structured extraction',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4.1-nano': {
      displayName: 'GPT-4.1 nano',
      description: 'Tiny GPT-4.1 option for classification, routing, and very high-volume tasks',
      lifecycle: 'deprecated',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1047576,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2024-04',
      structuredOutput: true,
      lastUpdated: '2025-04-14',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-4o': {
      displayName: 'GPT-4o',
      description: 'Omni-era GPT for multimodal chat, practical coding, and general assistants',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-08-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-mini': {
      displayName: 'GPT-4o mini',
      description: 'Small omni GPT for cheap multimodal assistance and production-scale traffic',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-07-18',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-4o-mini-search-preview': {
      displayName: 'GPT 4o Mini Search Preview',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      inputLimit: 111616,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2023-09',
      structuredOutput: false,
      lastUpdated: '2025-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-4o-mini-transcribe': {
      displayName: 'GPT-4o mini Transcribe',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2024-03-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'openai/gpt-4o-transcribe': {
      displayName: 'GPT-4o Transcribe',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2024-03-13',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'openai/gpt-5': {
      displayName: 'GPT-5',
      description: 'Original GPT-5 workhorse for reasoning, coding, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5-codex': {
      displayName: 'GPT-5-Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: true,
      lastUpdated: '2025-09-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5-mini': {
      displayName: 'GPT-5 Mini',
      description: 'Small GPT-5 for responsive agents, coding help, and everyday automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5-nano': {
      displayName: 'GPT-5 Nano',
      description: 'Tiny GPT-5 lane for routing, extraction, classification, and bulk jobs',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-05-30',
      structuredOutput: true,
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5-pro': {
      displayName: 'GPT-5 pro',
      description: 'Higher-accuracy GPT-5 tier for tough analysis, coding reviews, and planning',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 128000,
      maxOutputTokens: 272000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-10-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-codex': {
      displayName: 'GPT-5.1-Codex',
      description: 'Codex GPT for repository edits, code review, and practical software agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-codex-max': {
      displayName: 'GPT 5.1 Codex Max',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-codex-mini': {
      displayName: 'GPT-5.1 Codex mini',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-instant': {
      displayName: 'GPT-5.1 Instant',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      inputLimit: 111616,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-thinking': {
      displayName: 'GPT 5.1 Thinking',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text', 'image'] },
    },
    'openai/gpt-5.2': {
      displayName: 'GPT-5.2',
      description:
        'Reliable GPT generation for broad coding, writing, and tool-assisted product work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.2-codex': {
      displayName: 'GPT-5.2-Codex',
      description:
        'Code-specialist GPT for repository edits, reviews, and long-running software agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.2-pro': {
      displayName: 'GPT 5.2 ',
      description: 'Higher-accuracy GPT-5.2 variant for tougher reasoning and review workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2024-10',
      structuredOutput: false,
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.3-chat': {
      displayName: 'GPT-5.3 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      inputLimit: 111616,
      maxOutputTokens: 16384,
      lastUpdated: '2026-03-06',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.3-codex': {
      displayName: 'GPT 5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-02-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4': {
      displayName: 'GPT 5.4',
      description: 'Agent-ready GPT for coding and computer-use workflows at a lower cost',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4-mini': {
      displayName: 'GPT 5.4 Mini',
      description: 'Strong small GPT for coding subagents, quick tool use, and high-volume work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4-nano': {
      displayName: 'GPT 5.4 Nano',
      description: 'Cheapest GPT-5.4 lane for simple routing, extraction, and bulk automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 400000,
      inputLimit: 272000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: true,
      lastUpdated: '2026-03-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.4-pro': {
      displayName: 'GPT 5.4 Pro',
      description: 'More exact GPT-5.4 tier for demanding professional reasoning and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      structuredOutput: false,
      lastUpdated: '2026-03-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5': {
      displayName: 'GPT 5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      inputLimit: 872000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5-pro': {
      displayName: 'GPT 5.5 Pro',
      description: 'Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      inputLimit: 872000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-luna': {
      displayName: 'GPT 5.6 Luna',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-sol': {
      displayName: 'GPT 5.6 Sol',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-terra': {
      displayName: 'GPT 5.6 Terra',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-image-1': {
      displayName: 'GPT Image 1',
      description:
        'OpenAI image model for production generation, edits, and brand-safe visual workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-04-24',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'openai/gpt-image-1-mini': {
      displayName: 'GPT Image 1 Mini',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-10-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'openai/gpt-image-1.5': {
      displayName: 'GPT Image 1.5',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-11-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'openai/gpt-image-2': {
      displayName: 'GPT Image 2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'openai/gpt-oss-120b': {
      displayName: 'GPT OSS 120B',
      description: 'Open GPT reasoning model for self-hosted agents and controllable deployments',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-20b': {
      displayName: 'GPT OSS 20B',
      description:
        'Open-weight GPT model for self-hosted reasoning and instruction-following workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      inputLimit: 122880,
      maxOutputTokens: 8192,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-08-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-oss-safeguard-20b': {
      displayName: 'gpt-oss-safeguard-20b',
      description:
        'Safety model for policy screening, moderation, and risk-aware routing workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      inputLimit: 65536,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-12-01',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-realtime-1.5': {
      displayName: 'GPT-Realtime-1.5',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-realtime-2': {
      displayName: 'gpt-realtime-2',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-realtime-2.1': {
      displayName: 'gpt-realtime-2.1',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      inputLimit: 96000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2024-09-30',
      structuredOutput: false,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-realtime-mini': {
      displayName: 'GPT-Realtime mini',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2025-10-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'openai/gpt-realtime-whisper': {
      displayName: 'gpt-realtime-whisper',
      description:
        'Streaming speech-to-text model for low-latency transcript deltas from live audio',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-07',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'openai/o1': {
      displayName: 'o1',
      description: 'O-series reasoning model for hard analysis, math, coding, and planning',
      lifecycle: 'deprecated',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2023-09',
      structuredOutput: true,
      lastUpdated: '2024-12-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o3': {
      displayName: 'o3',
      description: 'Deliberate o-series reasoner for hard math, coding, and multi-step analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o3-deep-research': {
      displayName: 'o3-deep-research',
      description:
        'Research model for long-horizon investigation, synthesis, and analytical reports',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      inputLimit: 100000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2024-06-26',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o3-mini': {
      displayName: 'o3-mini',
      description: 'Smaller o-series reasoner for economical coding, math, and planning tasks',
      lifecycle: 'deprecated',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-01-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/o3-pro': {
      displayName: 'o3 Pro',
      description: 'High-effort o3 tier for difficult technical reasoning and careful answers',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      inputLimit: 100000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-10',
      structuredOutput: true,
      lastUpdated: '2025-06-10',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/o4-mini': {
      displayName: 'o4-mini',
      description: 'Fast o-series model for compact reasoning, coding, and tool use',
      lifecycle: 'deprecated',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 100000,
      knowledgeCutoff: '2024-05',
      structuredOutput: true,
      lastUpdated: '2025-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/text-embedding-3-large': {
      displayName: 'text-embedding-3-large',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      inputLimit: 6656,
      maxOutputTokens: 1536,
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/text-embedding-3-small': {
      displayName: 'text-embedding-3-small',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      inputLimit: 6656,
      maxOutputTokens: 1536,
      lastUpdated: '2024-01-25',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/text-embedding-ada-002': {
      displayName: 'text-embedding-ada-002',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      inputLimit: 6656,
      maxOutputTokens: 1536,
      lastUpdated: '2022-12-15',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/tts-1': {
      displayName: 'TTS-1',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2023-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'openai/tts-1-hd': {
      displayName: 'TTS-1 HD',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2023-11-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'openai/whisper-1': {
      displayName: 'Whisper',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2022-09-21',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'perplexity/pplx-embed-v1-0.6b': {
      displayName: 'Embed v1 0.6b',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-26',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'perplexity/pplx-embed-v1-4b': {
      displayName: 'Embed v1 4b',
      description:
        'Embedding model for semantic search, retrieval, clustering, and ranking pipelines',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-26',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'perplexity/sonar': {
      displayName: 'Sonar',
      description: 'Sonar search model for current answers, retrieval, and citation-backed chat',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 127000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2025-02',
      lastUpdated: '2025-02-19',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar-pro': {
      displayName: 'Sonar Pro',
      description: 'Advanced Sonar search model for deeper research and cited synthesis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2025-02-19',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'perplexity/sonar-reasoning-pro': {
      displayName: 'Sonar Reasoning Pro',
      description: 'Web-grounded reasoning model for multi-step research and cited answers',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 127000,
      maxOutputTokens: 8000,
      knowledgeCutoff: '2025-09',
      lastUpdated: '2025-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['minimal', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'poolside/laguna-s-2.1': {
      displayName: 'Laguna S 2.1',
      description: 'Agentic coding model from Poolside in the XS size class for local deployment',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: false,
      lastUpdated: '2026-07-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'poolside/laguna-s-2.1-free': {
      displayName: 'Laguna S 2.1 Free',
      description: 'Free provider route for experiments, demos, and cost-sensitive chat workloads',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 32768,
      structuredOutput: false,
      lastUpdated: '2026-07-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'prodia/flux-fast-schnell': {
      displayName: 'Flux Schnell',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2026-06-08',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'quiverai/arrow-1.1': {
      displayName: 'Arrow 1.1',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 131072,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v2': {
      displayName: 'Recraft V2',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2024-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v3': {
      displayName: 'Recraft V3',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 512,
      maxOutputTokens: 0,
      lastUpdated: '2024-10',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4': {
      displayName: 'Recraft V4',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-17',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4-pro': {
      displayName: 'Recraft V4 Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-17',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4.1': {
      displayName: 'Recraft V4.1',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-14',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4.1-pro': {
      displayName: 'Recraft V4.1 Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-14',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4.1-utility': {
      displayName: 'Recraft V4.1 Utility',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-14',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'recraft/recraft-v4.1-utility-pro': {
      displayName: 'Recraft V4.1 Utility Pro',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-14',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['image'] },
    },
    'sakana/fugu-ultra': {
      displayName: 'Fugu Ultra',
      description: 'Quality-first multi-agent model for hard research, analysis, and competitions',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      structuredOutput: true,
      lastUpdated: '2026-06-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'stepfun/step-3.5-flash': {
      displayName: 'StepFun 3.5 Flash',
      description: 'StepFun flash lane for quick multimodal reasoning and coding assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262114,
      maxOutputTokens: 262114,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'stepfun/step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'tencent/hy3': {
      displayName: 'Hy3',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      lastUpdated: '2026-07-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'thinkingmachines/inkling': {
      displayName: 'Inkling',
      description:
        'Multimodal MoE reasoning model (975B total, 41B active) for text, image, and audio',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2026-07-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'thinkingmachines/inkling-small': {
      displayName: 'Inkling Small',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      lastUpdated: '2026-07-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'minimal', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'voyage/rerank-2.5': {
      displayName: 'Voyage Rerank 2.5',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2025-08-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/rerank-2.5-lite': {
      displayName: 'Voyage Rerank 2.5 Lite',
      description:
        'Reranking model for improving retrieval quality in search and recommendation systems',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 32000,
      lastUpdated: '2025-08-11',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-3-large': {
      displayName: 'voyage-3-large',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-3.5': {
      displayName: 'voyage-3.5',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-3.5-lite': {
      displayName: 'voyage-3.5-lite',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2025-05-20',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-4': {
      displayName: 'voyage-4',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-4-large': {
      displayName: 'voyage-4-large',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-4-lite': {
      displayName: 'voyage-4-lite',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 32000,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-06',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-code-2': {
      displayName: 'voyage-code-2',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-01',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-code-3': {
      displayName: 'voyage-code-3',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-09',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-finance-2': {
      displayName: 'voyage-finance-2',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'voyage/voyage-law-2': {
      displayName: 'voyage-law-2',
      description: 'General-purpose chat model for instruction following, writing, and analysis',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 8192,
      maxOutputTokens: 1536,
      lastUpdated: '2024-03',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text'] },
    },
    'xai/grok-4.1-fast-non-reasoning': {
      displayName: 'Grok 4.1 Fast Non-Reasoning',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-07-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.1-fast-reasoning': {
      displayName: 'Grok 4.1 Fast Reasoning',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-multi-agent': {
      displayName: 'Grok 4.20 Multi-Agent',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-multi-agent-beta': {
      displayName: 'Grok 4.20 Multi Agent Beta',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-non-reasoning': {
      displayName: 'Grok 4.20 Non-Reasoning',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-23',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-non-reasoning-beta': {
      displayName: 'Grok 4.20 Beta Non-Reasoning',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-reasoning': {
      displayName: 'Grok 4.20 Reasoning',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.20-reasoning-beta': {
      displayName: 'Grok 4.20 Beta Reasoning',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 2000000,
      maxOutputTokens: 2000000,
      lastUpdated: '2026-03-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.3': {
      displayName: 'Grok 4.3',
      description:
        "xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-4.5': {
      displayName: 'Grok 4.5',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xai/grok-build-0.1': {
      displayName: 'Grok Build 0.1',
      description: 'Fast Grok coding model tuned for agentic engineering and iterative edits',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'xai/grok-imagine-image': {
      displayName: 'Grok Imagine Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-02-19',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['text', 'image'] },
    },
    'xai/grok-imagine-video': {
      displayName: 'Grok Imagine',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-01-28',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'xai/grok-imagine-video-1.5': {
      displayName: 'Grok Imagine Video 1.5',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-30',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'xai/grok-imagine-video-1.5-preview': {
      displayName: 'Grok Imagine Video 1.5 Preview',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-30',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: [] },
    },
    'xai/grok-stt': {
      displayName: 'Grok STT',
      description: 'Speech transcription model for accurate audio-to-text and captioning workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['audio'], output: ['text'] },
    },
    'xai/grok-tts': {
      displayName: 'Grok TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'xai/grok-voice-think-fast-1.0': {
      displayName: 'Grok Voice Think Fast 1.0',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-23',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'xai/grok-voice-think-fast-2.0': {
      displayName: 'Grok Voice Think Fast 2.0',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 0,
      maxOutputTokens: 0,
      lastUpdated: '2026-07-29',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'audio'], output: ['text', 'audio'] },
    },
    'xiaomi/mimo-v2.5': {
      displayName: 'MiMo M2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      maxOutputTokens: 131100,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'xiaomi/mimo-v2.5-pro': {
      displayName: 'MiMo V2.5 Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1050000,
      maxOutputTokens: 131000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.5': {
      displayName: 'GLM 4.5',
      description: 'Hybrid-reasoning GLM release that made the 4.5 line broadly useful',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 96000,
      knowledgeCutoff: '2025-07',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.5-air': {
      displayName: 'GLM 4.5 Air',
      description: 'Lighter GLM-4.5 variant for fast coding assistance and cheaper agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 96000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.5v': {
      displayName: 'GLM 4.5V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 66000,
      maxOutputTokens: 16000,
      knowledgeCutoff: '2025-08',
      lastUpdated: '2025-08-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'zai/glm-4.6': {
      displayName: 'GLM 4.6',
      description: 'Late GLM-4 workhorse for coding agents, reasoning, and structured tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 96000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.6v': {
      displayName: 'GLM-4.6V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 24000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'zai/glm-4.6v-flash': {
      displayName: 'GLM-4.6V-Flash',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 128000,
      maxOutputTokens: 24000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-09-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'zai/glm-4.7': {
      displayName: 'GLM 4.7',
      description: 'Mature GLM model for dependable coding, reasoning, and structured agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 120000,
      knowledgeCutoff: '2024-10',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.7-flash': {
      displayName: 'GLM 4.7 Flash',
      description: 'Budget GLM lane for fast coding help, routing, and everyday automation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 131000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-4.7-flashx': {
      displayName: 'GLM 4.7 FlashX',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-01',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 202800,
      maxOutputTokens: 131100,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5-turbo': {
      displayName: 'GLM 5 Turbo',
      description: 'Faster GLM-5 lane for coding agents that need lower latency',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 202800,
      maxOutputTokens: 131100,
      structuredOutput: true,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5.1': {
      displayName: 'GLM 5.1',
      description:
        'Strong GLM coding model for agentic engineering, terminals, and repository generation',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 202800,
      maxOutputTokens: 64000,
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5.2': {
      displayName: 'GLM 5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5.2-fast': {
      displayName: 'GLM 5.2 Fast',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'xhigh'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'zai/glm-5v-turbo': {
      displayName: 'GLM 5V Turbo',
      description: 'Fast GLM vision model for screenshots, documents, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl:
        'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
  },
  xai: {
    'grok-4.20-0309-non-reasoning': {
      displayName: 'Grok 4.20 (Non-Reasoning)',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1000000,
      maxOutputTokens: 30000,
      structuredOutput: true,
      lastUpdated: '2026-03-09',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.20-0309-reasoning': {
      displayName: 'Grok 4.20 (Reasoning)',
      description: 'Reasoning Grok for document-heavy analysis and long-horizon tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1000000,
      maxOutputTokens: 30000,
      structuredOutput: true,
      lastUpdated: '2026-03-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.20-multi-agent-0309': {
      displayName: 'Grok 4.20 Multi-Agent',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1000000,
      maxOutputTokens: 30000,
      structuredOutput: true,
      lastUpdated: '2026-03-09',
      capabilities: { vision: true, reasoning: true, functionCalling: false },
      thinkingOptions: { efforts: ['low', 'medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.3': {
      displayName: 'Grok 4.3',
      description: "xAI's Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1000000,
      maxOutputTokens: 30000,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-build-0.1': {
      displayName: 'Grok Build 0.1',
      description: 'Fast Grok coding model tuned for agentic engineering and iterative edits',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'grok-imagine-image': {
      displayName: 'Grok Imagine Image',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 8000,
      maxOutputTokens: 0,
      lastUpdated: '2026-01-28',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: ['image'] },
    },
    'grok-imagine-image-quality': {
      displayName: 'Grok Imagine Image Quality',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 8000,
      maxOutputTokens: 0,
      lastUpdated: '2026-04-03',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: ['image'] },
    },
    'grok-imagine-video': {
      displayName: 'Grok Imagine Video',
      description: 'Image model for prompt-driven generation, editing, and visual design workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1024,
      maxOutputTokens: 0,
      lastUpdated: '2026-01-28',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'pdf'], output: [] },
    },
    'grok-imagine-video-1.5': {
      displayName: 'Grok Imagine Video 1.5',
      description: 'Video model for image-to-video generation, editing, and extension workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.x.ai/docs/models',
      contextWindow: 1024,
      maxOutputTokens: 0,
      lastUpdated: '2026-05-30',
      capabilities: { vision: true, reasoning: false, functionCalling: false },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: [] },
    },
  },
  xiaomi: {
    'mimo-v2-flash': {
      displayName: 'MiMo-V2-Flash',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-12-01',
      lastUpdated: '2026-06-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-omni': {
      displayName: 'MiMo-V2-Omni',
      description: 'Legacy model retained for compatibility with older integrations',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 262144,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-06-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'mimo-v2-pro': {
      displayName: 'MiMo-V2-Pro',
      description: 'Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-06-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-06-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-06-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5-pro-ultraspeed': {
      displayName: 'MiMo-V2.5-Pro-UltraSpeed',
      description: 'MiMo pro model for strong multimodal reasoning and agent execution',
      lifecycle: 'beta',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-06-09',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  'xiaomi-token-plan-cn': {
    'mimo-v2-pro': {
      displayName: 'MiMo-V2-Pro',
      description: 'Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-tts': {
      displayName: 'MiMo-V2-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5-tts': {
      displayName: 'MiMo-V2.5-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voiceclone': {
      displayName: 'MiMo-V2.5-TTS-VoiceClone',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voicedesign': {
      displayName: 'MiMo-V2.5-TTS-VoiceDesign',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
  },
  'xiaomi-token-plan-sgp': {
    'mimo-v2-pro': {
      displayName: 'MiMo-V2-Pro',
      description: 'Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-tts': {
      displayName: 'MiMo-V2-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5-tts': {
      displayName: 'MiMo-V2.5-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voiceclone': {
      displayName: 'MiMo-V2.5-TTS-VoiceClone',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voicedesign': {
      displayName: 'MiMo-V2.5-TTS-VoiceDesign',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
  },
  'xiaomi-token-plan-ams': {
    'mimo-v2-pro': {
      displayName: 'MiMo-V2-Pro',
      description: 'Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks',
      lifecycle: 'deprecated',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2-tts': {
      displayName: 'MiMo-V2-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'mimo-v2.5-tts': {
      displayName: 'MiMo-V2.5-TTS',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voiceclone': {
      displayName: 'MiMo-V2.5-TTS-VoiceClone',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
    'mimo-v2.5-tts-voicedesign': {
      displayName: 'MiMo-V2.5-TTS-VoiceDesign',
      description: 'Speech generation model for controllable voice, narration, and audio delivery',
      lifecycle: 'active',
      docsUrl: 'https://platform.xiaomimimo.com/#/docs',
      contextWindow: 8192,
      maxOutputTokens: 8192,
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: false, functionCalling: false },
      modalities: { input: ['text'], output: ['audio'] },
    },
  },
  zai: {
    'glm-4.5': {
      displayName: 'GLM-4.5',
      description: 'Hybrid-reasoning GLM release that made the 4.5 line broadly useful',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.5-air': {
      displayName: 'GLM-4.5-Air',
      description: 'Lighter GLM-4.5 variant for fast coding assistance and cheaper agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.5-flash': {
      displayName: 'GLM-4.5-Flash',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 131072,
      maxOutputTokens: 98304,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-07-28',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.5v': {
      displayName: 'GLM-4.5V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 64000,
      maxOutputTokens: 16384,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-08-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'glm-4.6': {
      displayName: 'GLM-4.6',
      description: 'Late GLM-4 workhorse for coding agents, reasoning, and structured tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.6v': {
      displayName: 'GLM-4.6V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 128000,
      maxOutputTokens: 32768,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Mature GLM model for dependable coding, reasoning, and structured agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.7-flash': {
      displayName: 'GLM-4.7-Flash',
      description: 'Budget GLM lane for fast coding help, routing, and everyday automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-4.7-flashx': {
      displayName: 'GLM-4.7-FlashX',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5': {
      displayName: 'GLM-5',
      description:
        'General GLM flagship for coding, analysis, and tool-heavy engineering workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5-turbo': {
      displayName: 'GLM-5-Turbo',
      description: 'Faster GLM-5 lane for coding agents that need lower latency',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.1': {
      displayName: 'GLM-5.1',
      description:
        'Strong GLM coding model for agentic engineering, terminals, and repository generation',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-07',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5v-turbo': {
      displayName: 'GLM-5V-Turbo',
      description: 'Fast GLM vision model for screenshots, documents, and multimodal agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/guides/overview/pricing',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
  },
  'zai-coding-plan': {
    'glm-4.7': {
      displayName: 'GLM-4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/devpack/overview',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5-turbo': {
      displayName: 'GLM-5-Turbo',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/devpack/overview',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-03-16',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2': {
      displayName: 'GLM-5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/devpack/overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'glm-5.2-highspeed': {
      displayName: 'GLM-5.2 Highspeed',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.z.ai/devpack/overview',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
  },
  zenmux: {
    'anthropic/claude-3.5-haiku': {
      displayName: 'Claude 3.5 Haiku',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2024-11-04',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'anthropic/claude-3.7-sonnet': {
      displayName: 'Claude 3.7 Sonnet',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-02-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-fable-5': {
      displayName: 'Claude Fable 5',
      description: 'Claude model for creative writing, analysis, and controlled agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-haiku-4.5': {
      displayName: 'Claude Haiku 4.5',
      description:
        'Fast Claude model for responsive assistance, classification, and lightweight agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-10-15',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'anthropic/claude-opus-4': {
      displayName: 'Claude Opus 4',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.1': {
      displayName: 'Claude Opus 4.1',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-08-05',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.5': {
      displayName: 'Claude Opus 4.5',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-24',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'anthropic/claude-opus-4.6': {
      displayName: 'Claude Opus 4.6',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-05-31',
      lastUpdated: '2026-02-06',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'anthropic/claude-opus-4.7': {
      displayName: 'Claude Opus 4.7',
      description: 'Flagship Claude model for deep reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-opus-4.8': {
      displayName: 'Claude Opus 4.8',
      description:
        'Top Claude Opus tier for the hardest reasoning, coding, and long-horizon agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01',
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4': {
      displayName: 'Claude Sonnet 4',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-05-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.5': {
      displayName: 'Claude Sonnet 4.5',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-4.6': {
      displayName: 'Claude Sonnet 4.6',
      description: 'Balanced Claude model for coding, analysis, agent workflows, and cost control',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-02-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'anthropic/claude-sonnet-5': {
      displayName: 'Claude Sonnet 5',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'anthropic/claude-sonnet-5-free': {
      displayName: 'Claude Sonnet 5 (Free)',
      description: 'Everyday Claude agent model for coding, planning, browsing, and general work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-01-31',
      lastUpdated: '2026-06-30',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'baidu/ernie-5.0-thinking-preview': {
      displayName: 'ERNIE 5.0',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'deepseek/deepseek-chat': {
      displayName: 'DeepSeek-V3.2 (Non-thinking Mode)',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-01',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2': {
      displayName: 'DeepSeek V3.2',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-05',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v3.2-exp': {
      displayName: 'DeepSeek-V3.2-Exp',
      description: 'DeepSeek chat model for instruction following, coding, and analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 163000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-29',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-flash': {
      displayName: 'DeepSeek V4 Flash',
      description: 'Fast DeepSeek V4 lane for economical reasoning, coding, and long-context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'deepseek/deepseek-v4-pro': {
      displayName: 'DeepSeek V4 Pro',
      description: 'Open MoE flagship with million-token context for coding and long agent runs',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 384000,
      knowledgeCutoff: '2025-05',
      structuredOutput: true,
      lastUpdated: '2026-04-24',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'], toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'google/gemini-2.5-flash': {
      displayName: 'Gemini 2.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['pdf', 'image', 'text', 'audio'], output: ['text'] },
    },
    'google/gemini-2.5-flash-lite': {
      displayName: 'Gemini 2.5 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-22',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['pdf', 'image', 'text', 'audio'], output: ['text'] },
    },
    'google/gemini-2.5-pro': {
      displayName: 'Gemini 2.5 Pro',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-06-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['pdf', 'image', 'text', 'audio'], output: ['text'] },
    },
    'google/gemini-3-flash-preview': {
      displayName: 'Gemini 3 Flash Preview',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    'google/gemini-3.1-flash-lite': {
      displayName: 'Gemini 3.1 Flash Lite',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'google/gemini-3.1-flash-lite-preview': {
      displayName: 'Gemini 3.1 Flash Lite Preview',
      description: 'Low-latency Gemini model for high-volume multimodal and agent workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      maxOutputTokens: 65530,
      lastUpdated: '2025-03-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'google/gemini-3.1-pro-preview': {
      displayName: 'Gemini 3.1 Pro Preview',
      description: 'Advanced Gemini model for complex reasoning, coding, and multimodal analysis',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-02-19',
      lastUpdated: '2026-02-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf', 'audio'], output: ['text'] },
    },
    'google/gemini-3.5-flash': {
      displayName: 'Gemini 3.5 Flash',
      description: 'Fast Gemini model balancing multimodal reasoning, tool use, and cost',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-05-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'inclusionai/ling-1t': {
      displayName: 'Ling-1T',
      description:
        'Tool-capable chat model for instruction following and agentic application workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-10-09',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ring-1t': {
      displayName: 'Ring-1T',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-10-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'inclusionai/ring-2.6-1t': {
      displayName: 'inclusionAI: Ring-2.6-1T',
      description:
        'Reasoning model for deliberate analysis, multi-step problem solving, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262000,
      maxOutputTokens: 65000,
      knowledgeCutoff: '2025-12-31',
      lastUpdated: '2026-05-14',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'kuaishou/kat-coder-pro-v2': {
      displayName: 'KAT-Coder-Pro-V2',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 80000,
      lastUpdated: '2026-03-30',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2': {
      displayName: 'MiniMax M2',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-10-27',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.1': {
      displayName: 'MiniMax M2.1',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.5': {
      displayName: 'MiniMax M2.5',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.5-lightning': {
      displayName: 'MiniMax M2.5 highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204800,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-02-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.7': {
      displayName: 'MiniMax M2.7',
      description: 'MiniMax model for chat, coding, office work, and agentic tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204800,
      maxOutputTokens: 131070,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m2.7-highspeed': {
      displayName: 'MiniMax M2.7 highspeed',
      description: 'High-speed MiniMax model for low-latency coding and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 204800,
      maxOutputTokens: 131070,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'minimax/minimax-m3': {
      displayName: 'MiniMax-M3',
      description:
        'MiniMax multimodal model for long-context coding, perception, and agent planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 512000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-06-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2-0905': {
      displayName: 'Kimi K2 0905',
      description: 'Kimi model for long-context chat, coding, and agentic reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-04',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-thinking': {
      displayName: 'Kimi K2 Thinking',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2-thinking-turbo': {
      displayName: 'Kimi K2 Thinking Turbo',
      description: 'Kimi reasoning model for long-horizon research, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-06',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'moonshotai/kimi-k2.5': {
      displayName: 'Kimi K2.5',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-27',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.6': {
      displayName: 'Kimi K2.6',
      description: 'Kimi multimodal agent model for visual understanding, coding, and planning',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262140,
      maxOutputTokens: 262140,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-04-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.7-code': {
      displayName: 'Kimi K2.7 Code',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k2.7-code-free': {
      displayName: 'Kimi K2.7 Code (Free)',
      description:
        'Coding-focused Kimi model, stronger on long-horizon repo work with less overthinking',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262144,
      maxOutputTokens: 262144,
      knowledgeCutoff: '2025-01',
      structuredOutput: true,
      lastUpdated: '2026-06-12',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k3': {
      displayName: 'Kimi K3',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'moonshotai/kimi-k3-free': {
      displayName: 'Kimi K3 (Free)',
      description:
        'Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-07-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['max'], toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5': {
      displayName: 'GPT-5',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-08-07',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5-codex': {
      displayName: 'GPT-5 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.1': {
      displayName: 'GPT-5.1',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.1-chat': {
      displayName: 'GPT-5.1 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['pdf', 'image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.1-codex': {
      displayName: 'GPT-5.1-Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.1-codex-mini': {
      displayName: 'GPT-5.1-Codex-Mini',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-13',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'openai/gpt-5.2': {
      displayName: 'GPT-5.2',
      description: 'GPT model for general reasoning, writing, coding, and tool-assisted tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['image', 'text', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.2-codex': {
      displayName: 'GPT-5.2-Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-15',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.2-pro': {
      displayName: 'GPT-5.2-Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2025-12-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.3-chat': {
      displayName: 'GPT-5.3 Chat',
      description:
        'Chat-tuned GPT model for conversational assistance, writing, and tool workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 16380,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-5.3-codex': {
      displayName: 'GPT-5.3 Codex',
      description:
        'Coding-optimized GPT model for repository edits, reviews, and agentic software work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-5.4': {
      displayName: 'GPT-5.4',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.4-mini': {
      displayName: 'GPT-5.4 Mini',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-5.4-nano': {
      displayName: 'GPT-5.4 Nano',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'openai/gpt-5.4-pro': {
      displayName: 'GPT-5.4 Pro',
      description: 'Frontier GPT model for professional reasoning, coding, and multimodal work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'openai/gpt-5.5': {
      displayName: 'GPT-5.5',
      description: 'Default frontier GPT for coding, computer use, research, and knowledge work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5-instant': {
      displayName: 'GPT-5.5 Instant',
      description: 'Compact GPT model for low-latency assistance and high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 400000,
      inputLimit: 400000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-05-28',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.5-pro': {
      displayName: 'GPT-5.5 Pro',
      description: 'Highest-accuracy GPT-5.5 tier for slower, precision-heavy reasoning and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-12-01',
      structuredOutput: true,
      lastUpdated: '2026-04-23',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['medium', 'high', 'xhigh'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-luna': {
      displayName: 'GPT-5.6 Luna',
      description: 'Cost-efficient GPT-5.6 model for fast, high-volume workloads',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-sol': {
      displayName: 'GPT-5.6 Sol',
      description:
        'Frontier GPT-5.6 model for complex professional work, coding, and agentic workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'openai/gpt-5.6-terra': {
      displayName: 'GPT-5.6 Terra',
      description: 'Balanced GPT-5.6 model for capable, cost-efficient everyday work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1050000,
      inputLimit: 922000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2026-02-16',
      structuredOutput: true,
      lastUpdated: '2026-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high', 'xhigh', 'max'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'qwen/qwen3-coder-plus': {
      displayName: 'Qwen3-Coder-Plus',
      description: 'Qwen coding model for software agents, repository edits, and code reasoning',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-23',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3-max': {
      displayName: 'Qwen3-Max-Thinking',
      description: 'Qwen reasoning model for deliberate problem solving, math, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.5-flash': {
      displayName: 'Qwen3.5 Flash',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1020000,
      maxOutputTokens: 1020000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.5-plus': {
      displayName: 'Qwen3.5 Plus',
      description: 'Qwen vision-language model for visual reasoning, documents, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'qwen/qwen3.6-plus': {
      displayName: 'Qwen3.6-Plus',
      description: 'Qwen instruction model for multilingual chat, reasoning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-03-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.7-max': {
      displayName: 'Qwen3.7 Max',
      description:
        'Qwen frontier model tuned for agent frameworks, coding assistants, and long tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 65536,
      lastUpdated: '2026-05-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'qwen/qwen3.7-plus': {
      displayName: 'Qwen3.7 Plus',
      description: 'Multimodal Qwen workhorse for long-context agents, visual inputs, and coding',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-04',
      lastUpdated: '2026-06-02',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'sapiens-ai/agnes-1.5-lite': {
      displayName: 'Agnes 1.5 Lite',
      description: 'Efficient model for low-latency assistance, extraction, and routine automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2026-03-26',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'sapiens-ai/agnes-1.5-pro': {
      displayName: 'Agnes 1.5 Pro',
      description: 'Flagship model for demanding analysis, coding, and production agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      lastUpdated: '2026-03-21',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun/step-3': {
      displayName: 'Step-3',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 65536,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-31',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'stepfun/step-3.5-flash': {
      displayName: 'Step 3.5 Flash',
      description: 'StepFun flash model for efficient multimodal reasoning, coding, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-02-02',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'stepfun/step-3.7-flash': {
      displayName: 'Step 3.7 Flash',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'stepfun/step-3.7-flash-free': {
      displayName: 'Step 3.7 Flash (Free)',
      description: 'Newer StepFun flash model for faster agents, coding, and multimodal prompts',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      inputLimit: 256000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2026-03-01',
      lastUpdated: '2026-05-29',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'tencent/hy3-preview': {
      displayName: 'Hy3 preview',
      description: 'Tencent Hy reasoning model for coding, instruction following, and agent tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      lastUpdated: '2026-04-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'high'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'volcengine/doubao-seed-1.8': {
      displayName: 'Doubao-Seed-1.8',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'volcengine/doubao-seed-2.0-code': {
      displayName: 'Doubao Seed 2.0 Code',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 32000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: false, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'volcengine/doubao-seed-2.0-lite': {
      displayName: 'Doubao-Seed-2.0-lite',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-02-14',
      lastUpdated: '2026-02-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'volcengine/doubao-seed-2.0-mini': {
      displayName: 'Doubao-Seed-2.0-mini',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-02-14',
      lastUpdated: '2026-02-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'volcengine/doubao-seed-2.0-pro': {
      displayName: 'Doubao-Seed-2.0-pro',
      description: 'Multimodal reasoning model for visual analysis, planning, and tool use',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2026-02-14',
      lastUpdated: '2026-02-14',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'volcengine/doubao-seed-code': {
      displayName: 'Doubao-Seed-Code',
      description:
        'Coding model for repository understanding, refactors, and agentic engineering tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-11',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4': {
      displayName: 'Grok 4',
      description: 'Grok model for agentic tool use, reasoning, coding, and live assistance',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-09',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['image', 'text'], output: ['text'] },
    },
    'x-ai/grok-4-fast': {
      displayName: 'Grok 4 Fast',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 2000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-19',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4.1-fast': {
      displayName: 'Grok 4.1 Fast',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 2000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4.1-fast-non-reasoning': {
      displayName: 'Grok 4.1 Fast Non Reasoning',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 2000000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-11-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4.2-fast': {
      displayName: 'Grok 4.2 Fast',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 2000000,
      maxOutputTokens: 30000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4.2-fast-non-reasoning': {
      displayName: 'Grok 4.2 Fast Non Reasoning',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 2000000,
      maxOutputTokens: 30000,
      knowledgeCutoff: '2025-08-31',
      lastUpdated: '2026-03-20',
      capabilities: { vision: true, reasoning: false, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'x-ai/grok-4.3': {
      displayName: 'Grok 4.3',
      description:
        "xAI's default Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 1000000,
      structuredOutput: true,
      lastUpdated: '2026-04-17',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['none', 'low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-4.5': {
      displayName: 'Grok 4.5',
      description:
        "xAI's latest Grok for chat, coding, agentic tools, and lower hallucination risk",
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 500000,
      maxOutputTokens: 500000,
      structuredOutput: true,
      lastUpdated: '2026-07-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['low', 'medium', 'high'] },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-build-0.1': {
      displayName: 'Grok Build 0.1',
      description: 'Fast Grok coding model tuned for agentic engineering and iterative edits',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 256000,
      structuredOutput: true,
      lastUpdated: '2026-04-16',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
    'x-ai/grok-code-fast-1': {
      displayName: 'Grok Code Fast 1',
      description: 'Fast Grok model for responsive chat, reasoning, and tool-assisted work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 256000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-08-26',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'xiaomi/mimo-v2-flash': {
      displayName: 'MiMo-V2-Flash',
      description: 'MiMo flash model for fast multimodal assistance and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 262144,
      maxOutputTokens: 65536,
      knowledgeCutoff: '2024-12-01',
      lastUpdated: '2026-02-04',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'xiaomi/mimo-v2-omni': {
      displayName: 'MiMo V2 Omni',
      description: 'MiMo omni model for text, image, video, audio, and agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 265000,
      maxOutputTokens: 265000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio', 'pdf'], output: ['text'] },
    },
    'xiaomi/mimo-v2-pro': {
      displayName: 'MiMo V2 Pro',
      description: 'Earlier MiMo Pro model for multimodal agents, reasoning, and code tasks',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 256000,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-03-18',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'xiaomi/mimo-v2.5': {
      displayName: 'MiMo-V2.5',
      description: 'Open MiMo model for multimodal coding agents and long-context automation',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text', 'image', 'audio'], output: ['text'] },
    },
    'xiaomi/mimo-v2.5-pro': {
      displayName: 'MiMo-V2.5-Pro',
      description: 'Stronger MiMo Pro tier for multimodal reasoning and coding-agent execution',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1048576,
      maxOutputTokens: 131072,
      knowledgeCutoff: '2024-12',
      lastUpdated: '2026-04-22',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.5': {
      displayName: 'GLM 4.5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.5-air': {
      displayName: 'GLM 4.5 Air',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 128000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-07-25',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.6': {
      displayName: 'GLM 4.6',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-09-30',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.6v': {
      displayName: 'GLM 4.6V',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'z-ai/glm-4.6v-flash': {
      displayName: 'GLM 4.6V FlashX',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'z-ai/glm-4.6v-flash-free': {
      displayName: 'GLM 4.6V Flash (Free)',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-08',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image'], output: ['text'] },
    },
    'z-ai/glm-4.7': {
      displayName: 'GLM 4.7',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2025-12-23',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.7-flash-free': {
      displayName: 'GLM 4.7 Flash (Free)',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-4.7-flashx': {
      displayName: 'GLM 4.7 FlashX',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 64000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-01-19',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5': {
      displayName: 'GLM 5',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-02-12',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5-turbo': {
      displayName: 'GLM 5 Turbo',
      description: 'Efficient GLM model for fast reasoning, coding, and agent workflows',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      knowledgeCutoff: '2025-01-01',
      lastUpdated: '2026-03-20',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.1': {
      displayName: 'GLM-5.1',
      description: 'Flagship GLM model for hybrid reasoning, coding, and agentic engineering',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-04-03',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { toggle: true },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.2': {
      displayName: 'GLM 5.2',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5.2-free': {
      displayName: 'GLM 5.2 (Free)',
      description:
        'Open flagship GLM for long-horizon coding agents and million-token context work',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 1000000,
      maxOutputTokens: 131072,
      structuredOutput: true,
      lastUpdated: '2026-06-13',
      capabilities: { vision: false, reasoning: true, functionCalling: true },
      thinkingOptions: { efforts: ['high', 'max'] },
      modalities: { input: ['text'], output: ['text'] },
    },
    'z-ai/glm-5v-turbo': {
      displayName: 'GLM 5V Turbo',
      description: 'GLM vision model for visual reasoning, documents, and multimodal agents',
      lifecycle: 'active',
      docsUrl: 'https://docs.zenmux.ai',
      contextWindow: 200000,
      maxOutputTokens: 128000,
      lastUpdated: '2026-04-01',
      capabilities: { vision: true, reasoning: true, functionCalling: true },
      modalities: { input: ['text', 'image', 'pdf'], output: ['text'] },
    },
  },
};

export const GENERATED_MODELS_DEV_MODEL_PROVIDER_OVERRIDES: Record<
  | 'anthropic'
  | 'alibaba'
  | 'alibaba-coding-plan-cn'
  | 'alibaba-coding-plan'
  | 'alibaba-token-plan-cn'
  | 'alibaba-token-plan'
  | 'cerebras'
  | 'cohere'
  | 'cloudflare-workers-ai'
  | 'deepinfra'
  | 'deepseek'
  | 'fireworks-ai'
  | 'github-copilot'
  | 'google'
  | 'gemini-cli'
  | 'groq'
  | 'huggingface'
  | 'kimi-coding-plan'
  | 'MiniMax'
  | 'MiniMax-cn'
  | 'minimax-coding-plan'
  | 'mistral'
  | 'moonshot'
  | 'nvidia'
  | 'ollama-cloud'
  | 'openai'
  | 'opencode'
  | 'opencode-go'
  | 'openrouter'
  | 'siliconflow'
  | 'stepfun'
  | 'stepfun-ai'
  | 'stepfun-ai-step-plan'
  | 'stepfun-step-plan'
  | 'togetherai'
  | 'tencent-coding-plan'
  | 'tencent-token-plan'
  | 'tencent-tokenhub'
  | 'vercel'
  | 'xai'
  | 'xiaomi'
  | 'xiaomi-token-plan-cn'
  | 'xiaomi-token-plan-sgp'
  | 'xiaomi-token-plan-ams'
  | 'zai'
  | 'zai-coding-plan'
  | 'zenmux',
  Record<string, { npm: string; api?: string }>
> = {
  anthropic: {},
  alibaba: {},
  'alibaba-coding-plan-cn': {},
  'alibaba-coding-plan': {},
  'alibaba-token-plan-cn': {},
  'alibaba-token-plan': {},
  cerebras: {},
  cohere: {
    'north-mini-code-1-0': {
      npm: '@ai-sdk/openai-compatible',
      api: 'https://api.cohere.ai/compatibility/v1',
    },
  },
  'cloudflare-workers-ai': {},
  deepinfra: {},
  deepseek: {},
  'fireworks-ai': {},
  'github-copilot': {},
  google: {},
  'gemini-cli': {},
  groq: {},
  huggingface: {},
  'kimi-coding-plan': {},
  MiniMax: {},
  'MiniMax-cn': {},
  'minimax-coding-plan': {},
  mistral: {},
  moonshot: {},
  nvidia: {},
  'ollama-cloud': {},
  openai: {},
  opencode: {
    'claude-3-5-haiku': { npm: '@ai-sdk/anthropic' },
    'claude-fable-5': { npm: '@ai-sdk/anthropic' },
    'claude-haiku-4-5': { npm: '@ai-sdk/anthropic' },
    'claude-opus-4-1': { npm: '@ai-sdk/anthropic' },
    'claude-opus-4-5': { npm: '@ai-sdk/anthropic' },
    'claude-opus-4-6': { npm: '@ai-sdk/anthropic' },
    'claude-opus-4-7': { npm: '@ai-sdk/anthropic' },
    'claude-opus-4-8': { npm: '@ai-sdk/anthropic' },
    'claude-opus-5': { npm: '@ai-sdk/anthropic' },
    'claude-sonnet-4': { npm: '@ai-sdk/anthropic' },
    'claude-sonnet-4-5': { npm: '@ai-sdk/anthropic' },
    'claude-sonnet-4-6': { npm: '@ai-sdk/anthropic' },
    'claude-sonnet-5': { npm: '@ai-sdk/anthropic' },
    'gemini-3-flash': { npm: '@ai-sdk/google' },
    'gemini-3-pro': { npm: '@ai-sdk/google' },
    'gemini-3.1-pro': { npm: '@ai-sdk/google' },
    'gemini-3.5-flash': { npm: '@ai-sdk/google' },
    'gemini-3.5-flash-lite': { npm: '@ai-sdk/google' },
    'gemini-3.6-flash': { npm: '@ai-sdk/google' },
    'gpt-5': { npm: '@ai-sdk/openai' },
    'gpt-5-codex': { npm: '@ai-sdk/openai' },
    'gpt-5-nano': { npm: '@ai-sdk/openai' },
    'gpt-5.1': { npm: '@ai-sdk/openai' },
    'gpt-5.1-codex': { npm: '@ai-sdk/openai' },
    'gpt-5.1-codex-max': { npm: '@ai-sdk/openai' },
    'gpt-5.1-codex-mini': { npm: '@ai-sdk/openai' },
    'gpt-5.2': { npm: '@ai-sdk/openai' },
    'gpt-5.2-codex': { npm: '@ai-sdk/openai' },
    'gpt-5.3-codex': { npm: '@ai-sdk/openai' },
    'gpt-5.3-codex-spark': { npm: '@ai-sdk/openai' },
    'gpt-5.4': { npm: '@ai-sdk/openai' },
    'gpt-5.4-mini': { npm: '@ai-sdk/openai' },
    'gpt-5.4-nano': { npm: '@ai-sdk/openai' },
    'gpt-5.4-pro': { npm: '@ai-sdk/openai' },
    'gpt-5.5': { npm: '@ai-sdk/openai' },
    'gpt-5.5-pro': { npm: '@ai-sdk/openai' },
    'gpt-5.6-luna': { npm: '@ai-sdk/openai' },
    'gpt-5.6-sol': { npm: '@ai-sdk/openai' },
    'gpt-5.6-terra': { npm: '@ai-sdk/openai' },
    'grok-4.5': { npm: '@ai-sdk/openai' },
    'minimax-m2.1-free': { npm: '@ai-sdk/anthropic' },
    'minimax-m2.5-free': { npm: '@ai-sdk/anthropic' },
    'minimax-m3-free': { npm: '@ai-sdk/anthropic' },
    'qwen3.5-plus': { npm: '@ai-sdk/anthropic' },
    'qwen3.6-plus': { npm: '@ai-sdk/anthropic' },
    'qwen3.6-plus-free': { npm: '@ai-sdk/anthropic' },
  },
  'opencode-go': {
    'gpt-5.6-luna': { npm: '@ai-sdk/openai' },
    'grok-4.5': { npm: '@ai-sdk/openai' },
    'minimax-m2.5': { npm: '@ai-sdk/anthropic' },
    'minimax-m2.7': { npm: '@ai-sdk/anthropic' },
    'minimax-m3': { npm: '@ai-sdk/anthropic' },
    'qwen3.5-plus': { npm: '@ai-sdk/anthropic' },
    'qwen3.6-plus': { npm: '@ai-sdk/anthropic' },
    'qwen3.7-max': { npm: '@ai-sdk/anthropic' },
    'qwen3.7-plus': { npm: '@ai-sdk/anthropic' },
    'qwen3.8-max': { npm: '@ai-sdk/anthropic' },
  },
  openrouter: {},
  siliconflow: {},
  stepfun: {},
  'stepfun-ai': {},
  'stepfun-ai-step-plan': {},
  'stepfun-step-plan': {},
  togetherai: {},
  'tencent-coding-plan': {},
  'tencent-token-plan': {},
  'tencent-tokenhub': {},
  vercel: {},
  xai: {},
  xiaomi: {},
  'xiaomi-token-plan-cn': {},
  'xiaomi-token-plan-sgp': {},
  'xiaomi-token-plan-ams': {},
  zai: {},
  'zai-coding-plan': {},
  zenmux: {
    'anthropic/claude-3.5-haiku': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-3.7-sonnet': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-fable-5': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-haiku-4.5': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4.1': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4.5': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4.6': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4.7': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-opus-4.8': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-sonnet-4': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-sonnet-4.5': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-sonnet-4.6': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-sonnet-5': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'anthropic/claude-sonnet-5-free': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'minimax/minimax-m2': { npm: '@ai-sdk/anthropic', api: 'https://zenmux.ai/api/anthropic/v1' },
    'minimax/minimax-m2.1': { npm: '@ai-sdk/anthropic', api: 'https://zenmux.ai/api/anthropic/v1' },
    'minimax/minimax-m2.5': { npm: '@ai-sdk/anthropic', api: 'https://zenmux.ai/api/anthropic/v1' },
    'minimax/minimax-m2.5-lightning': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'minimax/minimax-m2.7': { npm: '@ai-sdk/anthropic', api: 'https://zenmux.ai/api/anthropic/v1' },
    'minimax/minimax-m2.7-highspeed': {
      npm: '@ai-sdk/anthropic',
      api: 'https://zenmux.ai/api/anthropic/v1',
    },
    'minimax/minimax-m3': { npm: '@ai-sdk/anthropic', api: 'https://zenmux.ai/api/anthropic/v1' },
    'openai/gpt-5': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5-codex': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.1': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.1-chat': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.1-codex': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.1-codex-mini': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.2': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.2-codex': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.2-pro': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.3-chat': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.3-codex': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.4': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.4-mini': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.4-nano': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
    'openai/gpt-5.4-pro': { npm: '@ai-sdk/openai', api: 'https://zenmux.ai/api/v1' },
  },
};

export const GENERATED_MODELS_DEV_PROVIDER_FACTS: Record<
  | 'anthropic'
  | 'alibaba'
  | 'alibaba-coding-plan-cn'
  | 'alibaba-coding-plan'
  | 'alibaba-token-plan-cn'
  | 'alibaba-token-plan'
  | 'cerebras'
  | 'cohere'
  | 'cloudflare-workers-ai'
  | 'deepinfra'
  | 'deepseek'
  | 'fireworks-ai'
  | 'github-copilot'
  | 'google'
  | 'gemini-cli'
  | 'groq'
  | 'huggingface'
  | 'kimi-coding-plan'
  | 'MiniMax'
  | 'MiniMax-cn'
  | 'minimax-coding-plan'
  | 'mistral'
  | 'moonshot'
  | 'nvidia'
  | 'ollama-cloud'
  | 'openai'
  | 'opencode'
  | 'opencode-go'
  | 'openrouter'
  | 'siliconflow'
  | 'stepfun'
  | 'stepfun-ai'
  | 'stepfun-ai-step-plan'
  | 'stepfun-step-plan'
  | 'togetherai'
  | 'tencent-coding-plan'
  | 'tencent-token-plan'
  | 'tencent-tokenhub'
  | 'vercel'
  | 'xai'
  | 'xiaomi'
  | 'xiaomi-token-plan-cn'
  | 'xiaomi-token-plan-sgp'
  | 'xiaomi-token-plan-ams'
  | 'zai'
  | 'zai-coding-plan'
  | 'zenmux',
  { id: string; name: string; api?: string; doc: string }
> = {
  anthropic: {
    id: 'anthropic',
    name: 'Anthropic',
    doc: 'https://docs.anthropic.com/en/docs/about-claude/models',
  },
  alibaba: {
    id: 'alibaba',
    name: 'Alibaba',
    api: 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1',
    doc: 'https://www.alibabacloud.com/help/en/model-studio/models',
  },
  'alibaba-coding-plan-cn': {
    id: 'alibaba-coding-plan-cn',
    name: 'Alibaba Coding Plan (China)',
    api: 'https://coding.dashscope.aliyuncs.com/v1',
    doc: 'https://help.aliyun.com/zh/model-studio/coding-plan',
  },
  'alibaba-coding-plan': {
    id: 'alibaba-coding-plan',
    name: 'Alibaba Coding Plan',
    api: 'https://coding-intl.dashscope.aliyuncs.com/v1',
    doc: 'https://www.alibabacloud.com/help/en/model-studio/coding-plan',
  },
  'alibaba-token-plan-cn': {
    id: 'alibaba-token-plan-cn',
    name: 'Alibaba Token Plan (China)',
    api: 'https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1',
    doc: 'https://www.alibabacloud.com/help/zh/model-studio/token-plan-overview',
  },
  'alibaba-token-plan': {
    id: 'alibaba-token-plan',
    name: 'Alibaba Token Plan',
    api: 'https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1',
    doc: 'https://www.alibabacloud.com/help/en/model-studio/token-plan-overview',
  },
  cerebras: {
    id: 'cerebras',
    name: 'Cerebras',
    doc: 'https://inference-docs.cerebras.ai/models/overview',
  },
  cohere: { id: 'cohere', name: 'Cohere', doc: 'https://docs.cohere.com/docs/models' },
  'cloudflare-workers-ai': {
    id: 'cloudflare-workers-ai',
    name: 'Cloudflare Workers AI',
    api: 'https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/ai/v1',
    doc: 'https://developers.cloudflare.com/workers-ai/models/',
  },
  deepinfra: { id: 'deepinfra', name: 'Deep Infra', doc: 'https://deepinfra.com/models' },
  deepseek: {
    id: 'deepseek',
    name: 'DeepSeek',
    api: 'https://api.deepseek.com',
    doc: 'https://api-docs.deepseek.com/quick_start/pricing',
  },
  'fireworks-ai': {
    id: 'fireworks-ai',
    name: 'Fireworks AI',
    api: 'https://api.fireworks.ai/inference/v1/',
    doc: 'https://fireworks.ai/docs/',
  },
  'github-copilot': {
    id: 'github-copilot',
    name: 'GitHub Copilot',
    api: 'https://api.githubcopilot.com',
    doc: 'https://docs.github.com/en/copilot',
  },
  google: { id: 'google', name: 'Google', doc: 'https://ai.google.dev/gemini-api/docs/models' },
  'gemini-cli': {
    id: 'google',
    name: 'Google',
    doc: 'https://ai.google.dev/gemini-api/docs/models',
  },
  groq: { id: 'groq', name: 'Groq', doc: 'https://console.groq.com/docs/models' },
  huggingface: {
    id: 'huggingface',
    name: 'Hugging Face',
    api: 'https://router.huggingface.co/v1',
    doc: 'https://huggingface.co/docs/inference-providers',
  },
  'kimi-coding-plan': {
    id: 'kimi-for-coding',
    name: 'Kimi For Coding',
    api: 'https://api.kimi.com/coding/v1',
    doc: 'https://www.kimi.com/code/docs/en/third-party-tools/other-coding-agents.html',
  },
  MiniMax: {
    id: 'minimax',
    name: 'MiniMax (minimax.io)',
    api: 'https://api.minimax.io/anthropic/v1',
    doc: 'https://platform.minimax.io/docs/guides/quickstart',
  },
  'MiniMax-cn': {
    id: 'minimax-cn',
    name: 'MiniMax (minimaxi.com)',
    api: 'https://api.minimaxi.com/anthropic/v1',
    doc: 'https://platform.minimaxi.com/docs/guides/quickstart',
  },
  'minimax-coding-plan': {
    id: 'minimax-coding-plan',
    name: 'MiniMax Token Plan (minimax.io)',
    api: 'https://api.minimax.io/anthropic/v1',
    doc: 'https://platform.minimax.io/docs/token-plan/intro',
  },
  mistral: {
    id: 'mistral',
    name: 'Mistral',
    doc: 'https://docs.mistral.ai/getting-started/models/',
  },
  moonshot: {
    id: 'moonshotai-cn',
    name: 'Moonshot AI (China)',
    api: 'https://api.moonshot.cn/v1',
    doc: 'https://platform.moonshot.cn/docs/api/chat',
  },
  nvidia: {
    id: 'nvidia',
    name: 'Nvidia',
    api: 'https://integrate.api.nvidia.com/v1',
    doc: 'https://docs.api.nvidia.com/nim/',
  },
  'ollama-cloud': {
    id: 'ollama-cloud',
    name: 'Ollama Cloud',
    api: 'https://ollama.com/v1',
    doc: 'https://docs.ollama.com/cloud',
  },
  openai: { id: 'openai', name: 'OpenAI', doc: 'https://platform.openai.com/docs/models' },
  opencode: {
    id: 'opencode',
    name: 'OpenCode Zen',
    api: 'https://opencode.ai/zen/v1',
    doc: 'https://opencode.ai/docs/zen',
  },
  'opencode-go': {
    id: 'opencode-go',
    name: 'OpenCode Go',
    api: 'https://opencode.ai/zen/go/v1',
    doc: 'https://opencode.ai/docs/zen',
  },
  openrouter: {
    id: 'openrouter',
    name: 'OpenRouter',
    api: 'https://openrouter.ai/api/v1',
    doc: 'https://openrouter.ai/models',
  },
  siliconflow: {
    id: 'siliconflow',
    name: 'SiliconFlow',
    api: 'https://api.siliconflow.com/v1',
    doc: 'https://cloud.siliconflow.com/models',
  },
  stepfun: {
    id: 'stepfun',
    name: 'StepFun (China)',
    api: 'https://api.stepfun.com/v1',
    doc: 'https://platform.stepfun.com/docs/zh/overview/concept',
  },
  'stepfun-ai': {
    id: 'stepfun-ai',
    name: 'StepFun (Global)',
    api: 'https://api.stepfun.ai/v1',
    doc: 'https://platform.stepfun.ai/docs/en/overview/concept',
  },
  'stepfun-ai-step-plan': {
    id: 'stepfun-ai-step-plan',
    name: 'StepFun Step Plan (Global)',
    api: 'https://api.stepfun.ai/step_plan/v1',
    doc: 'https://platform.stepfun.ai/docs/en/step-plan/integrations/reasoning-api',
  },
  'stepfun-step-plan': {
    id: 'stepfun-step-plan',
    name: 'StepFun Step Plan (China)',
    api: 'https://api.stepfun.com/step_plan/v1',
    doc: 'https://platform.stepfun.com/docs/zh/step-plan/integrations/reasoning-api',
  },
  togetherai: {
    id: 'togetherai',
    name: 'Together AI',
    doc: 'https://docs.together.ai/docs/serverless-models',
  },
  'tencent-coding-plan': {
    id: 'tencent-coding-plan',
    name: 'Tencent Coding Plan (China)',
    api: 'https://api.lkeap.cloud.tencent.com/coding/v3',
    doc: 'https://cloud.tencent.com/document/product/1772/128947',
  },
  'tencent-token-plan': {
    id: 'tencent-token-plan',
    name: 'Tencent Token Plan',
    api: 'https://api.lkeap.cloud.tencent.com/plan/v3',
    doc: 'https://cloud.tencent.com/document/product/1823/130060',
  },
  'tencent-tokenhub': {
    id: 'tencent-tokenhub',
    name: 'Tencent TokenHub',
    api: 'https://tokenhub.tencentmaas.com/v1',
    doc: 'https://cloud.tencent.com/document/product/1823/130050',
  },
  vercel: {
    id: 'vercel',
    name: 'Vercel AI Gateway',
    doc: 'https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway',
  },
  xai: { id: 'xai', name: 'xAI', doc: 'https://docs.x.ai/docs/models' },
  xiaomi: {
    id: 'xiaomi',
    name: 'Xiaomi',
    api: 'https://api.xiaomimimo.com/v1',
    doc: 'https://platform.xiaomimimo.com/#/docs',
  },
  'xiaomi-token-plan-cn': {
    id: 'xiaomi-token-plan-cn',
    name: 'Xiaomi Token Plan (China)',
    api: 'https://token-plan-cn.xiaomimimo.com/v1',
    doc: 'https://platform.xiaomimimo.com/#/docs',
  },
  'xiaomi-token-plan-sgp': {
    id: 'xiaomi-token-plan-sgp',
    name: 'Xiaomi Token Plan (Singapore)',
    api: 'https://token-plan-sgp.xiaomimimo.com/v1',
    doc: 'https://platform.xiaomimimo.com/#/docs',
  },
  'xiaomi-token-plan-ams': {
    id: 'xiaomi-token-plan-ams',
    name: 'Xiaomi Token Plan (Europe)',
    api: 'https://token-plan-ams.xiaomimimo.com/v1',
    doc: 'https://platform.xiaomimimo.com/#/docs',
  },
  zai: {
    id: 'zai',
    name: 'Z.AI',
    api: 'https://api.z.ai/api/paas/v4',
    doc: 'https://docs.z.ai/guides/overview/pricing',
  },
  'zai-coding-plan': {
    id: 'zai-coding-plan',
    name: 'Z.AI Coding Plan',
    api: 'https://api.z.ai/api/coding/paas/v4',
    doc: 'https://docs.z.ai/devpack/overview',
  },
  zenmux: {
    id: 'zenmux',
    name: 'ZenMux',
    api: 'https://zenmux.ai/api/v1',
    doc: 'https://docs.zenmux.ai',
  },
};

export const GENERATED_MODELS_DEV_DIRECTORY: Record<string, { api?: string }> = {
  '302ai': { api: 'https://api.302.ai/v1' },
  abacus: { api: 'https://routellm.abacus.ai/v1' },
  'abliteration-ai': { api: 'https://api.abliteration.ai/v1' },
  'ai-router': { api: 'https://api.ai-router.dev/v1' },
  aiand: { api: 'https://api.aiand.com/v1' },
  aihubmix: {},
  'aki-io': { api: 'https://aki.io/v1' },
  alibaba: { api: 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1' },
  'alibaba-cn': { api: 'https://dashscope.aliyuncs.com/compatible-mode/v1' },
  'alibaba-coding-plan': { api: 'https://coding-intl.dashscope.aliyuncs.com/v1' },
  'alibaba-coding-plan-cn': { api: 'https://coding.dashscope.aliyuncs.com/v1' },
  'alibaba-token-plan': {
    api: 'https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1',
  },
  'alibaba-token-plan-cn': {
    api: 'https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1',
  },
  'amazon-bedrock': {},
  ambient: { api: 'https://api.ambient.xyz/v1' },
  anthropic: {},
  anyapi: { api: 'https://api.anyapi.ai/v1' },
  'atomic-chat': { api: 'http://127.0.0.1:1337/v1' },
  auriko: { api: 'https://api.auriko.ai/v1' },
  azure: {},
  'azure-cognitive-services': {},
  bailing: { api: 'https://api.tbox.cn/api/llm/v1/chat/completions' },
  baseten: { api: 'https://inference.baseten.co/v1' },
  berget: { api: 'https://api.berget.ai/v1' },
  blueclaw: { api: 'https://openai.blueclaw.network/v1' },
  cerebras: {},
  chutes: { api: 'https://llm.chutes.ai/v1' },
  clarifai: { api: 'https://api.clarifai.com/v2/ext/openai/v1' },
  claudinio: { api: 'https://api.claudin.io/v1' },
  'cline-pass': { api: 'https://api.cline.bot/api/v1' },
  'cloudferro-sherlock': { api: 'https://api-sherlock.cloudferro.com/openai/v1/' },
  'cloudflare-ai-gateway': {},
  'cloudflare-workers-ai': {
    api: 'https://api.cloudflare.com/client/v4/accounts/${CLOUDFLARE_ACCOUNT_ID}/ai/v1',
  },
  cohere: {},
  cortecs: { api: 'https://api.cortecs.ai/v1' },
  crof: { api: 'https://crof.ai/v1' },
  crossmodel: { api: 'https://api.crossmodel.ai/v1' },
  daoxe: { api: 'https://daoxe.com/v1' },
  databricks: { api: 'https://${DATABRICKS_HOST}/ai-gateway/mlflow/v1' },
  deepinfra: {},
  deepseek: { api: 'https://api.deepseek.com' },
  digitalocean: { api: 'https://inference.do-ai.run/v1' },
  dinference: { api: 'https://api.dinference.com/v1' },
  drun: { api: 'https://chat.d.run/v1' },
  ebcloud: { api: 'https://maas-api.ebcloud.com/v1' },
  empiriolabs: { api: 'https://api.empiriolabs.ai/v1' },
  evroc: { api: 'https://models.think.evroc.com/v1' },
  fastrouter: { api: 'https://go.fastrouter.ai/api/v1' },
  'fireworks-ai': { api: 'https://api.fireworks.ai/inference/v1/' },
  freemodel: { api: 'https://cc.freemodel.dev/v1' },
  friendli: { api: 'https://api.friendli.ai/serverless/v1' },
  frogbot: { api: 'https://app.frogbot.ai/api/v1' },
  'github-copilot': { api: 'https://api.githubcopilot.com' },
  gitlab: {},
  gmicloud: { api: 'https://api.gmi-serving.com/v1' },
  google: {},
  'google-vertex': {},
  'google-vertex-anthropic': {},
  greenpt: { api: 'https://api.greenpt.ai/v1' },
  groq: {},
  helicone: { api: 'https://ai-gateway.helicone.ai/v1' },
  hetzner: { api: 'https://inference.hetzner.com/api/v1' },
  'hpc-ai': { api: 'https://api.hpc-ai.com/inference/v1' },
  huggingface: { api: 'https://router.huggingface.co/v1' },
  hyper: { api: 'https://hyper.charm.land/v1' },
  iflowcn: { api: 'https://apis.iflow.cn/v1' },
  impossibl: { api: 'https://api.impossibl.com/v1' },
  inception: { api: 'https://api.inceptionlabs.ai/v1/' },
  inceptron: { api: 'https://api.inceptron.io/v1' },
  inference: { api: 'https://inference.net/v1' },
  inferx: { api: 'https://model.inferx.net/endpoints/v1' },
  infomaniak: { api: 'https://api.infomaniak.com/2/ai/${INFOMANIAK_PRODUCT_ID}/openai/v1' },
  'io-net': { api: 'https://api.intelligence.io.solutions/api/v1' },
  jiekou: { api: 'https://api.jiekou.ai/openai' },
  kenari: { api: 'https://kenari.id/v1' },
  kilo: { api: 'https://api.kilo.ai/api/gateway' },
  'kimi-for-coding': { api: 'https://api.kimi.com/coding/v1' },
  'kuae-cloud-coding-plan': { api: 'https://coding-plan-endpoint.kuaecloud.net/v1' },
  lilac: { api: 'https://api.getlilac.com/v1' },
  llama: { api: 'https://api.llama.com/compat/v1/' },
  llmgateway: { api: 'https://api.llmgateway.io/v1' },
  llmtr: { api: 'https://llmtr.com/v1' },
  lmstudio: { api: 'http://127.0.0.1:1234/v1' },
  longcat: { api: 'https://api.longcat.chat/openai' },
  lucidquery: { api: 'https://api.lucidquery.com/v1' },
  lynkr: { api: 'http://127.0.0.1:8081/v1' },
  meganova: { api: 'https://api.meganova.ai/v1' },
  'merge-gateway': {},
  meta: { api: 'https://api.meta.ai/v1' },
  minimax: { api: 'https://api.minimax.io/anthropic/v1' },
  'minimax-cn': { api: 'https://api.minimaxi.com/anthropic/v1' },
  'minimax-cn-coding-plan': { api: 'https://api.minimaxi.com/anthropic/v1' },
  'minimax-coding-plan': { api: 'https://api.minimax.io/anthropic/v1' },
  mistral: {},
  mixlayer: { api: 'https://models.mixlayer.ai/v1' },
  moark: { api: 'https://moark.com/v1' },
  modal: { api: 'https://inference.us-west.modal.direct/v1' },
  'model-oracle-ai': { api: 'https://api.modeloracle.com/api/v1' },
  modelis: { api: 'https://modelishub.com/v1' },
  modelscope: { api: 'https://api-inference.modelscope.cn/v1' },
  moonshotai: { api: 'https://api.moonshot.ai/v1' },
  'moonshotai-cn': { api: 'https://api.moonshot.cn/v1' },
  morph: { api: 'https://api.morphllm.com/v1' },
  'nano-gpt': { api: 'https://nano-gpt.com/api/v1' },
  nearai: { api: 'https://cloud-api.near.ai/v1' },
  nebius: { api: 'https://api.tokenfactory.nebius.com/v1' },
  neon: { api: '${NEON_AI_GATEWAY_BASE_URL}/v1' },
  neuralwatt: { api: 'https://api.neuralwatt.com/v1' },
  nova: { api: 'https://api.nova.amazon.com/v1' },
  'novita-ai': { api: 'https://api.novita.ai/openai' },
  nvidia: { api: 'https://integrate.api.nvidia.com/v1' },
  ofox: { api: 'https://api.ofox.ai/v1' },
  'ollama-cloud': { api: 'https://ollama.com/v1' },
  openai: {},
  opencode: { api: 'https://opencode.ai/zen/v1' },
  'opencode-go': { api: 'https://opencode.ai/zen/go/v1' },
  openrouter: { api: 'https://openrouter.ai/api/v1' },
  orcarouter: { api: 'https://api.orcarouter.ai/v1' },
  ovhcloud: { api: 'https://oai.endpoints.kepler.ai.cloud.ovh.net/v1' },
  perplexity: {},
  'perplexity-agent': { api: 'https://api.perplexity.ai/v1' },
  pioneer: { api: 'https://api.pioneer.ai/v1' },
  poe: { api: 'https://api.poe.com/v1' },
  poolside: { api: 'https://inference.poolside.ai/v1' },
  'privatemode-ai': { api: 'http://localhost:8080/v1' },
  'qihang-ai': { api: 'https://api.qhaigc.net/v1' },
  'qiniu-ai': { api: 'https://api.qnaigc.com/v1' },
  qvac: {},
  'regolo-ai': { api: 'https://api.regolo.ai/v1' },
  requesty: { api: 'https://router.requesty.ai/v1' },
  'routing-run': { api: 'https://api.routing.run/v1' },
  sakana: { api: 'https://api.sakana.ai/v1' },
  'sap-ai-core': {},
  sarvam: { api: 'https://api.sarvam.ai/v1' },
  scaleway: { api: 'https://api.scaleway.ai/v1' },
  scx: { api: 'https://api.scx.ai/v1' },
  siliconflow: { api: 'https://api.siliconflow.com/v1' },
  'siliconflow-cn': { api: 'https://api.siliconflow.cn/v1' },
  'snowflake-cortex': {
    api: 'https://${SNOWFLAKE_ACCOUNT}.snowflakecomputing.com/api/v2/cortex/v1',
  },
  stackit: { api: 'https://api.openai-compat.model-serving.eu01.onstackit.cloud/v1' },
  stepfun: { api: 'https://api.stepfun.com/v1' },
  'stepfun-ai': { api: 'https://api.stepfun.ai/v1' },
  'stepfun-ai-step-plan': { api: 'https://api.stepfun.ai/step_plan/v1' },
  'stepfun-step-plan': { api: 'https://api.stepfun.com/step_plan/v1' },
  subconscious: { api: 'https://api.subconscious.dev/v1' },
  submodel: { api: 'https://llm.submodel.ai/v1' },
  synthetic: { api: 'https://api.synthetic.new/openai/v1' },
  'tencent-coding-plan': { api: 'https://api.lkeap.cloud.tencent.com/coding/v3' },
  'tencent-token-plan': { api: 'https://api.lkeap.cloud.tencent.com/plan/v3' },
  'tencent-tokenhub': { api: 'https://tokenhub.tencentmaas.com/v1' },
  tensorx: { api: 'https://api.tensorx.ai/v1' },
  'the-grid-ai': { api: 'https://api.thegrid.ai/v1' },
  thinkingmachines: {
    api: 'https://tinker.thinkingmachines.dev/services/tinker-prod/anthropic/api/v1',
  },
  tinfoil: { api: 'https://inference.tinfoil.sh/v1' },
  togetherai: {},
  trustedrouter: { api: 'https://api.trustedrouter.com/v1' },
  'umans-ai': { api: 'https://api.code.umans.ai/v1' },
  'umans-ai-coding-plan': { api: 'https://api.code.umans.ai/v1' },
  unorouter: { api: 'https://api.unorouter.com/v1' },
  upstage: { api: 'https://api.upstage.ai/v1/solar' },
  v0: {},
  venice: {},
  vercel: {},
  vivgrid: { api: 'https://api.vivgrid.com/v1' },
  vultr: { api: 'https://api.vultrinference.com/v1' },
  'wafer.ai': { api: 'https://pass.wafer.ai/v1' },
  wandb: { api: 'https://api.inference.wandb.ai/v1' },
  xai: {},
  xiaomi: { api: 'https://api.xiaomimimo.com/v1' },
  'xiaomi-token-plan-ams': { api: 'https://token-plan-ams.xiaomimimo.com/v1' },
  'xiaomi-token-plan-cn': { api: 'https://token-plan-cn.xiaomimimo.com/v1' },
  'xiaomi-token-plan-sgp': { api: 'https://token-plan-sgp.xiaomimimo.com/v1' },
  xpersona: { api: 'https://www.xpersona.co/v1' },
  zai: { api: 'https://api.z.ai/api/paas/v4' },
  'zai-coding-plan': { api: 'https://api.z.ai/api/coding/paas/v4' },
  zeldoc: { api: 'https://api.zeldoc.ai/v1' },
  zenifra: { api: 'https://ai.zenifra.com/v1' },
  zenmux: { api: 'https://zenmux.ai/api/v1' },
  zhipuai: { api: 'https://open.bigmodel.cn/api/paas/v4' },
  'zhipuai-coding-plan': { api: 'https://open.bigmodel.cn/api/coding/paas/v4' },
};