{
 "meta": {
  "synced": "2026-10-06",
  "count": 45,
  "total_catalog": 189,
  "vendors": [
   "Anthropic",
   "ElevenLabs",
   "Google",
   "OpenAI",
   "xAI"
  ],
  "site": "https://models.cmdspace.work",
  "source": "CMDSPACE vault · 40. Docs/49. API Information (api-model-sync)"
 },
 "models": [
  {
   "api_name": "antigravity-preview-09-2026",
   "model_name": "Antigravity Agent",
   "model_family": "Gemini",
   "model_version": "preview-09-2026",
   "provider": "Google",
   "description": "Google's managed Antigravity coding agent on the Gemini Interactions API (released 2026-09-17), running in a Google-hosted Linux sandbox on Gemini 3.8 Flash. Replaces antigravity-preview-05-2026. Reference when calling Antigravity programmatically or migrating from the 05-2026 harness.",
   "release_date": "2026-09-17",
   "status": "preview",
   "context_window": 1048576,
   "max_output_tokens": 65536,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/interactions",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "agent",
    "coding",
    "sandbox",
    "antigravity"
   ],
   "aliases": [
    "antigravity-preview",
    "Antigravity Agent 09-2026"
   ],
   "supports": [
    "code_execution",
    "function_calling",
    "streaming",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Antigravity preview model page",
     "url": "https://ai.google.dev/gemini-api/docs/models/antigravity-preview-09-2026"
    },
    {
     "label": "Antigravity agent guide",
     "url": "https://ai.google.dev/gemini-api/docs/antigravity-agent"
    },
    {
     "label": "Gemini API changelog (2026-09-17)",
     "url": "https://ai.google.dev/gemini-api/docs/changelog"
    }
   ],
   "note_modified": "2026-09-28",
   "category": "agent"
  },
  {
   "api_name": "claude-fable-5-1",
   "model_name": "Claude Fable 5.1",
   "model_family": "Claude",
   "model_version": "5.1",
   "provider": "Anthropic",
   "description": "Anthropic's latest and most capable model (released 2026-09-01), extending Claude Fable 5 at identical input/output prices but with cache reads cut to a quarter (2.5% of input). Built for demanding reasoning and long-horizon agentic work. Reference when choosing between Opus 5 and the Fable tier, or when migrating off Fable 5 (three breaking changes).",
   "release_date": "2026-09-01",
   "knowledge_cutoff": "2026-06",
   "training_data_cutoff": "2026-06",
   "status": "active",
   "retirement_date": "2027-09-01",
   "context_window": 1000000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 10,
   "cached_input_price_per_1m": 0.25,
   "output_price_per_1m": 50,
   "batch_input_price_per_1m": 5,
   "batch_output_price_per_1m": 25,
   "default_effort": "high",
   "api_endpoint": "https://api.anthropic.com/v1/messages",
   "aws_bedrock_id": "anthropic.claude-fable-5-1",
   "gcp_vertex_id": "claude-fable-5-1",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "agentic",
    "flagship"
   ],
   "aliases": [
    "claude-fable-5-1",
    "Claude Fable 5.1",
    "Fable 5.1"
   ],
   "supports": [
    "adaptive_thinking",
    "batch_api",
    "function_calling",
    "prompt_caching",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Claude Fable 5.1 Overview",
     "url": "https://platform.claude.com/docs/en/models/fable-5-1/overview"
    },
    {
     "label": "What's new in Claude Fable 5.1",
     "url": "https://platform.claude.com/docs/en/models/fable-5-1/whats-new-fable-5-1"
    },
    {
     "label": "Models overview",
     "url": "https://platform.claude.com/docs/en/docs/about-claude/models/overview"
    }
   ],
   "note_modified": "2026-09-10",
   "category": "text"
  },
  {
   "api_name": "claude-haiku-4-5-20251001",
   "model_name": "Claude Haiku 4.5",
   "model_family": "Claude",
   "model_version": "4.5",
   "provider": "Anthropic",
   "description": "Fastest model with near-frontier intelligence",
   "release_date": "2025-10-01",
   "knowledge_cutoff": "2025-02",
   "training_data_cutoff": "2025-07",
   "status": "active",
   "context_window": 200000,
   "max_output_tokens": 64000,
   "input_price_per_1m": 1,
   "cached_input_price_per_1m": 0.1,
   "output_price_per_1m": 5,
   "api_endpoint": "https://api.anthropic.com/v1/messages",
   "aws_bedrock_id": "anthropic.claude-haiku-4-5-20251001-v1:0",
   "gcp_vertex_id": "claude-haiku-4-5@20251001",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "fast",
    "cost-efficient"
   ],
   "aliases": [
    "claude-haiku-4-5"
   ],
   "supports": [
    "batch_api",
    "extended_thinking",
    "function_calling",
    "priority_tier",
    "prompt_caching",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Anthropic models overview",
     "url": "https://platform.claude.com/docs/en/docs/about-claude/models/overview"
    }
   ],
   "note_modified": "2026-01-14",
   "category": "text"
  },
  {
   "api_name": "claude-opus-5-5",
   "model_name": "Claude Opus 5.5",
   "model_family": "Claude",
   "model_version": "5.5",
   "provider": "Anthropic",
   "description": "Anthropic's recommended default model (released 2026-09-22) for long-running agentic coding and knowledge work, with 1M context, 128K output, always-on adaptive thinking at medium default effort, and $4/$20 pricing. Reference for current Claude Opus specs, pricing, and the four breaking changes when migrating from Opus 5.",
   "release_date": "2026-09-22",
   "knowledge_cutoff": "2026-06",
   "training_data_cutoff": "2026-06",
   "status": "active",
   "retirement_not_before": "2027-09-22",
   "context_window": 1000000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 4.0,
   "cached_input_price_per_1m": 0.2,
   "output_price_per_1m": 20.0,
   "default_effort": "medium",
   "api_endpoint": "https://api.anthropic.com/v1/messages",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "agentic"
   ],
   "aliases": [
    "claude-opus-5-5",
    "Claude Opus 5.5",
    "Opus 5.5"
   ],
   "supports": [
    "batch_api",
    "function_calling",
    "prompt_caching",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Claude Opus 5.5 overview",
     "url": "https://platform.claude.com/docs/en/models/opus-5-5/overview"
    },
    {
     "label": "Models overview",
     "url": "https://platform.claude.com/docs/en/models/overview"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "text"
  },
  {
   "api_name": "claude-sonnet-5-5",
   "model_name": "Claude Sonnet 5.5",
   "model_family": "Claude",
   "model_version": "5.5",
   "provider": "Anthropic",
   "description": "Anthropic's current Sonnet-tier model (released 2026-09-28), the best combination of speed and intelligence, succeeding Sonnet 5. 1M context, 128K output, adaptive thinking with a new between_tools setting, $2/$10 per MTok and a Jun 2026 knowledge cutoff. Reference for balanced production Claude API usage and for the five breaking changes when migrating from Sonnet 5.",
   "release_date": "2026-09-28",
   "knowledge_cutoff": "2026-06",
   "training_data_cutoff": "2026-06",
   "status": "active",
   "retirement_not_before": "2027-09-28",
   "context_window": 1000000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 2,
   "cached_input_price_per_1m": 0.2,
   "output_price_per_1m": 10,
   "batch_input_price_per_1m": 1.0,
   "batch_output_price_per_1m": 5.0,
   "default_effort": "high",
   "api_endpoint": "https://api.anthropic.com/v1/messages",
   "aws_bedrock_id": "anthropic.claude-sonnet-5-5",
   "gcp_vertex_id": "claude-sonnet-5-5",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "vision",
    "agents"
   ],
   "aliases": [
    "claude-sonnet-5-5",
    "Claude Sonnet 5.5",
    "Sonnet 5.5"
   ],
   "supports": [
    "adaptive_thinking",
    "batch_api",
    "function_calling",
    "prompt_caching",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Models Overview",
     "url": "https://platform.claude.com/docs/en/about-claude/models/overview"
    },
    {
     "label": "Claude Sonnet 5.5 model page",
     "url": "https://platform.claude.com/docs/en/models/sonnet-5-5/overview"
    },
    {
     "label": "What's new in Claude Sonnet 5.5",
     "url": "https://platform.claude.com/docs/en/models/sonnet-5-5/whats-new-sonnet-5-5"
    },
    {
     "label": "Claude API Pricing",
     "url": "https://platform.claude.com/docs/en/about-claude/pricing"
    }
   ],
   "note_modified": "2026-10-01",
   "category": "text"
  },
  {
   "api_name": "deep-research-max-preview-04-2026",
   "model_name": "Gemini Deep Research Max",
   "model_family": "Gemini",
   "model_version": "max-preview-04-2026",
   "provider": "Google",
   "description": "Higher-capability variant of Google's Gemini Deep Research agent — deeper multi-step web research and synthesis. Preview; per-token pricing not published. Reference for the most thorough Gemini agentic research tier.",
   "release_date": "2026-04-01",
   "knowledge_cutoff": "2025-01",
   "status": "preview",
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "agent",
    "deep-research",
    "search-grounding"
   ],
   "aliases": [
    "deep-research-max-preview"
   ],
   "supports": [
    "function_calling",
    "search_grounding",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API Models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-06-20",
   "category": "agent"
  },
  {
   "api_name": "eleven_flash_v2_5",
   "model_name": "Eleven Flash v2.5",
   "model_family": "Eleven Flash",
   "model_version": "v2.5",
   "provider": "ElevenLabs",
   "description": "ElevenLabs ultra-low-latency Text-to-Speech model generating speech in ~75ms across 32 languages. Reference when building real-time conversational AI, voice agents, or live applications where latency matters more than expressive range.",
   "release_date": "2024-11-01",
   "status": "active",
   "context_window": 40000,
   "max_output_tokens": 0,
   "price_per_1k_characters": 0.05,
   "api_endpoint": "https://api.elevenlabs.io/v1/text-to-speech",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "tts",
    "audio",
    "speech",
    "voice",
    "elevenlabs",
    "low-latency",
    "realtime"
   ],
   "aliases": [
    "flash-v2-5",
    "elevenlabs-flash-v2-5"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "ElevenLabs models",
     "url": "https://elevenlabs.io/docs/overview/models"
    }
   ],
   "note_modified": "2026-04-29",
   "category": "speech"
  },
  {
   "api_name": "eleven_v3",
   "model_name": "Eleven v3",
   "model_family": "Eleven v3",
   "model_version": "v3",
   "provider": "ElevenLabs",
   "description": "ElevenLabs flagship Text-to-Speech model with the most expressive emotional range across 70+ languages. Supports inline audio tags ([whispers], [laughs], [excited]) and multi-speaker dialogue. Reference when generating dramatic voiceovers, audiobooks, or long-form expressive content.",
   "release_date": "2025-06-01",
   "status": "active",
   "context_window": 5000,
   "max_output_tokens": 0,
   "price_per_1k_characters": 0.1,
   "api_endpoint": "https://api.elevenlabs.io/v1/text-to-speech",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "tts",
    "audio",
    "speech",
    "voice",
    "elevenlabs"
   ],
   "aliases": [
    "eleven-v3",
    "elevenlabs-v3"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "ElevenLabs models",
     "url": "https://elevenlabs.io/docs/overview/models"
    }
   ],
   "note_modified": "2026-04-29",
   "category": "speech"
  },
  {
   "api_name": "gemini-3-pro-image",
   "model_name": "Gemini 3 Pro Image",
   "model_family": "Gemini",
   "model_version": "3",
   "provider": "Google",
   "description": "Google's GA flagship image generation model (Nano Banana Pro) based on Gemini 3 Pro — highest-fidelity image output with up to 4K resolution. Reference for premium image generation on Gemini and its per-image pricing.",
   "release_date": "2025-12-01",
   "knowledge_cutoff": "2025-01",
   "status": "active",
   "context_window": 65536,
   "max_output_tokens": 32768,
   "input_price_per_1m": 2.0,
   "price_per_image_1024": 0.134,
   "price_per_image_2k": 0.134,
   "price_per_image_4k": 0.24,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3-pro-image",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text",
    "image"
   ],
   "tags": [
    "image-generation",
    "multimodal",
    "creative",
    "4k"
   ],
   "aliases": [
    "nano-banana-pro",
    "gemini-3-pro-image"
   ],
   "supports": [
    "batch_api",
    "caching",
    "image_generation",
    "streaming",
    "structured_outputs"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-07-13",
   "category": "image"
  },
  {
   "api_name": "gemini-3.1-flash-image",
   "model_name": "Gemini 3.1 Flash Image",
   "model_family": "Gemini",
   "model_version": "3.1",
   "provider": "Google",
   "description": "Google's GA high-efficiency image generation model (Nano Banana 2) with up to 4K resolution, ~90% text rendering accuracy, and conversational editing. Reference for Flash-tier image generation on Gemini and its per-image pricing.",
   "release_date": "2026-02-26",
   "knowledge_cutoff": "2025-01",
   "status": "active",
   "context_window": 65536,
   "max_output_tokens": 65536,
   "input_price_per_1m": 0.5,
   "price_per_image_1024": 0.067,
   "price_per_image_2k": 0.101,
   "price_per_image_4k": 0.151,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.1-flash-image",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text",
    "image"
   ],
   "tags": [
    "image-generation",
    "multimodal",
    "text-rendering",
    "conversational-editing",
    "4k"
   ],
   "aliases": [
    "nano-banana-2",
    "gemini-3.1-flash-image"
   ],
   "supports": [
    "batch_api",
    "caching",
    "function_calling",
    "image_generation",
    "search_grounding",
    "streaming",
    "structured_outputs"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-07-13",
   "category": "image"
  },
  {
   "api_name": "gemini-3.1-pro-preview",
   "model_name": "Gemini 3.1 Pro",
   "model_family": "Gemini",
   "model_version": "3.1",
   "provider": "Google",
   "description": "Google's most intelligent model with frontier reasoning, multimodal input, and 1M context window — successor to Gemini 3 Pro Preview",
   "release_date": "2026-02-19",
   "knowledge_cutoff": "2025-01",
   "status": "preview",
   "context_window": 1048576,
   "max_output_tokens": 65536,
   "input_price_per_1m": 2.0,
   "cached_input_price_per_1m": 0.2,
   "output_price_per_1m": 12.0,
   "input_price_per_1m_long": 4.0,
   "output_price_per_1m_long": 18.0,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.1-pro-preview",
   "input_modalities": [
    "text",
    "image",
    "video",
    "audio",
    "pdf"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "multimodal",
    "reasoning",
    "agentic",
    "flagship"
   ],
   "aliases": [
    "gemini-3.1-pro",
    "gemini-3-1-pro-preview"
   ],
   "supports": [
    "batch_api",
    "caching",
    "code_execution",
    "file_search",
    "function_calling",
    "search_grounding",
    "streaming",
    "structured_outputs",
    "thinking",
    "url_context"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini 3.1 Pro Preview - Artificial Analysis",
     "url": "https://artificialanalysis.ai/models/gemini-3-1-pro-preview"
    },
    {
     "label": "Google Gemini Models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-03-20",
   "category": "text"
  },
  {
   "api_name": "gemini-3.5-flash-lite",
   "model_name": "Gemini 3.5 Flash-Lite",
   "model_family": "Gemini",
   "model_version": "3.5",
   "provider": "Google",
   "description": "Google's most cost-efficient GA model, optimized for high-volume agentic tasks. $0.30 input / $2.50 output per 1M tokens with multimodal input. Reference for budget-tier Gemini API usage and high-volume pipelines.",
   "release_date": "2026",
   "status": "active",
   "context_window": 1048576,
   "max_output_tokens": 65536,
   "input_price_per_1m": 0.3,
   "cached_input_price_per_1m": 0.03,
   "output_price_per_1m": 2.5,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.5-flash-lite:generateContent",
   "input_modalities": [
    "text",
    "image",
    "video",
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "multimodal",
    "budget"
   ],
   "aliases": [
    "gemini-3.5-flash-lite",
    "Gemini 3.5 Flash-Lite"
   ],
   "supports": [
    "batch_api",
    "function_calling",
    "prompt_caching",
    "streaming",
    "structured_outputs",
    "thinking",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API Models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-07-30",
   "category": "text"
  },
  {
   "api_name": "gemini-3.5-transcribe-live",
   "model_name": "Gemini 3.5 Transcribe Live",
   "model_family": "Gemini",
   "model_version": "3.5",
   "provider": "Google",
   "description": "Google's real-time streaming speech-to-text model, served through the Gemini Live API for low-latency live transcription with diarization. Reference when building live captioning, real-time meeting transcription, or comparing streaming STT cost against OpenAI gpt-live-transcribe.",
   "release_date": "2026-08",
   "status": "active",
   "context_window": 1048576,
   "input_price_per_1m": 3.5,
   "output_price_per_1m": 21.0,
   "audio_input_price_per_minute": 0.005,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/live",
   "input_modalities": [
    "audio",
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "stt",
    "speech-to-text",
    "transcription",
    "realtime",
    "streaming",
    "audio"
   ],
   "aliases": [
    "gemini-3.5-transcribe-live",
    "Gemini Transcribe Live"
   ],
   "supports": [
    "diarization",
    "realtime",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-08-31",
   "category": "transcribe"
  },
  {
   "api_name": "gemini-3.5-transcribe",
   "model_name": "Gemini 3.5 Transcribe",
   "model_family": "Gemini",
   "model_version": "3.5",
   "provider": "Google",
   "description": "Google's dedicated Gemini speech-to-text model — batch/unary transcription with speaker diarization, word-level timestamps, custom vocabulary, and 85+ locale detection via the Interactions API. Reference when choosing a Gemini STT model or comparing per-minute transcription cost against OpenAI gpt-transcribe or ElevenLabs Scribe.",
   "release_date": "2026-08",
   "status": "active",
   "context_window": 1048576,
   "input_price_per_1m": 2.0,
   "output_price_per_1m": 12.0,
   "audio_input_price_per_minute": 0.003,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/interactions",
   "input_modalities": [
    "audio",
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "stt",
    "speech-to-text",
    "transcription",
    "audio",
    "diarization"
   ],
   "aliases": [
    "gemini-3.5-transcribe",
    "Gemini Transcribe"
   ],
   "supports": [
    "custom_vocabulary",
    "diarization",
    "word_timestamps"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-08-31",
   "category": "transcribe"
  },
  {
   "api_name": "gemini-3.8-flash-lite-tts",
   "model_name": "Gemini 3.8 Flash-Lite TTS",
   "model_family": "Gemini",
   "model_version": "3.8",
   "provider": "Google",
   "description": "Google's high-throughput, low-latency text-to-speech model (GA 2026-09-22) that replaces gemini-3.1-flash-tts-preview. Reference for cost-efficient Gemini TTS in dubbing, voice-agent cascades, and read-aloud pipelines.",
   "release_date": "2026-09-22",
   "status": "active",
   "input_price_per_1m": 0.5,
   "batch_input_price_per_1m": 0.25,
   "input_price_per_1m_from_2027": 1.0,
   "audio_output_price_per_1m_from_2027": 12.0,
   "audio_output_price_per_1m": 6.0,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash-lite-tts",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "tts",
    "text-to-speech",
    "audio",
    "low-latency"
   ],
   "aliases": [
    "Gemini 3.8 Flash-Lite TTS",
    "gemini-flash-lite-tts"
   ],
   "supports": [
    "audio",
    "audio_generation",
    "batch_api",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    },
    {
     "label": "Gemini 3.8 TTS model page",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash-tts"
    },
    {
     "label": "Gemini API changelog (2026-09-22)",
     "url": "https://ai.google.dev/gemini-api/docs/changelog"
    }
   ],
   "note_modified": "2026-09-28",
   "category": "speech"
  },
  {
   "api_name": "gemini-3.8-flash-tts",
   "model_name": "Gemini 3.8 Flash TTS",
   "model_family": "Gemini",
   "model_version": "3.8",
   "provider": "Google",
   "description": "Google's flagship creative text-to-speech model (GA 2026-09-22) with voice design, consented voice replication, 130+ languages, and long-form multi-speaker stability. Reference for top-tier Gemini TTS and its promotional audio-token pricing that doubles on 2027-01-01.",
   "release_date": "2026-09-22",
   "status": "active",
   "context_window": 8192,
   "max_output_tokens": 16384,
   "input_price_per_1m": 0.5,
   "batch_input_price_per_1m": 0.25,
   "input_price_per_1m_from_2027": 1.0,
   "audio_output_price_per_1m_from_2027": 18.0,
   "audio_output_price_per_1m": 9.0,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash-tts",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "tts",
    "text-to-speech",
    "audio",
    "voice-design"
   ],
   "aliases": [
    "Gemini 3.8 Flash TTS",
    "gemini-flash-tts"
   ],
   "supports": [
    "audio",
    "audio_generation",
    "batch_api",
    "caching",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini 3.8 Flash TTS model page",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash-tts"
    },
    {
     "label": "Gemini API pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    },
    {
     "label": "Gemini API changelog (2026-09-22)",
     "url": "https://ai.google.dev/gemini-api/docs/changelog"
    },
    {
     "label": "Google blog: Gemini 3.8 TTS",
     "url": "https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-8-text-to-speech/"
    }
   ],
   "note_modified": "2026-09-28",
   "category": "speech"
  },
  {
   "api_name": "gemini-3.8-flash",
   "model_name": "Gemini 3.8 Flash",
   "model_family": "Gemini",
   "model_version": "3.8",
   "provider": "Google",
   "description": "Google's most intelligent Flash model (stable, September 2026), succeeding Gemini 3.7 Flash. Engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows. Price-matched to 3.7 Flash at $0.75/$3.75 per MTok through 2026, doubling in 2027. Reference for the current default Gemini API workhorse.",
   "release_date": "2026-09",
   "status": "active",
   "context_window": 1048576,
   "max_output_tokens": 65536,
   "input_price_per_1m": 0.75,
   "cached_input_price_per_1m": 0.075,
   "output_price_per_1m": 3.75,
   "batch_input_price_per_1m": 0.375,
   "batch_output_price_per_1m": 1.875,
   "promotional_pricing_until": "2026-12-31",
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/models/gemini-3.8-flash:generateContent",
   "input_modalities": [
    "text",
    "image",
    "video",
    "audio",
    "pdf"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "multimodal",
    "agents"
   ],
   "aliases": [
    "gemini-3.8-flash",
    "Gemini 3.8 Flash"
   ],
   "supports": [
    "audio",
    "batch_api",
    "code_execution",
    "computer_use",
    "function_calling",
    "prompt_caching",
    "search_grounding",
    "streaming",
    "structured_outputs",
    "thinking",
    "video",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini 3.8 Flash Model Page",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-05",
   "category": "text"
  },
  {
   "api_name": "gemini-3.8-live-extended-thinking",
   "model_name": "Gemini 3.8 Live Extended Thinking",
   "model_family": "Gemini",
   "model_version": "3.8",
   "provider": "Google",
   "description": "Google's stable Live API model (Sep 2026) for heavier reasoning during real-time voice conversations; reasoning and tool calls run in the background while audio keeps streaming. Same limits and pricing as Gemini 3.8 Live. Reference for complex multi-step Gemini voice agents.",
   "release_date": "2026-09",
   "status": "active",
   "context_window": 131072,
   "max_output_tokens": 65536,
   "text_input_price_per_1m": 0.75,
   "text_output_price_per_1m": 4.5,
   "audio_input_price_per_1m": 3.0,
   "audio_output_price_per_1m": 12.0,
   "audio_input_price_per_minute": 0.005,
   "audio_output_price_per_minute": 0.018,
   "image_input_price_per_1m": 1.0,
   "api_endpoint": "wss://generativelanguage.googleapis.com/ws/google.ai.generativelanguage.v1beta.GenerativeService.BidiGenerateContent",
   "input_modalities": [
    "text",
    "image",
    "audio",
    "video"
   ],
   "output_modalities": [
    "text",
    "audio"
   ],
   "tags": [
    "voice",
    "realtime",
    "live-api",
    "audio"
   ],
   "aliases": [
    "gemini-3.8-live-extended-thinking",
    "Gemini 3.8 Live Extended Thinking"
   ],
   "supports": [
    "audio",
    "function_calling",
    "reasoning",
    "streaming",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini 3.8 Live Extended Thinking model page",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-3.8-live-extended-thinking"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "realtime"
  },
  {
   "api_name": "gemini-3.8-live",
   "model_name": "Gemini 3.8 Live",
   "model_family": "Gemini",
   "model_version": "3.8",
   "provider": "Google",
   "description": "Google's stable default Live API model (Sep 2026) for low-latency voice agents, replacing gemini-3.1-flash-live-preview. Audio-to-audio with text/image/video input, 131K input and 65K output tokens, interleaved thinking, async function calling. Reference for Gemini real-time voice agents.",
   "release_date": "2026-09",
   "status": "active",
   "context_window": 131072,
   "max_output_tokens": 65536,
   "text_input_price_per_1m": 0.75,
   "text_output_price_per_1m": 4.5,
   "audio_input_price_per_1m": 3.0,
   "audio_output_price_per_1m": 12.0,
   "audio_input_price_per_minute": 0.005,
   "audio_output_price_per_minute": 0.018,
   "image_input_price_per_1m": 1.0,
   "api_endpoint": "wss://generativelanguage.googleapis.com/ws/google.ai.generativelanguage.v1beta.GenerativeService.BidiGenerateContent",
   "input_modalities": [
    "text",
    "image",
    "audio",
    "video"
   ],
   "output_modalities": [
    "text",
    "audio"
   ],
   "tags": [
    "voice",
    "realtime",
    "live-api",
    "audio"
   ],
   "aliases": [
    "gemini-3.8-live",
    "Gemini 3.8 Live"
   ],
   "supports": [
    "audio",
    "function_calling",
    "reasoning",
    "streaming",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini 3.8 Live model page",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-3.8-live"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "realtime"
  },
  {
   "api_name": "gemini-embedding-2-preview",
   "model_name": "Gemini Embedding 2",
   "model_family": "Gemini",
   "model_version": "2",
   "provider": "Google",
   "description": "Google's latest multimodal embedding model supporting text, image, audio, and video inputs. Reference for current Gemini embedding pricing and modality coverage for RAG and semantic search.",
   "release_date": "2026-05-19",
   "status": "preview",
   "text_input_price_per_1m": 0.2,
   "audio_input_price_per_1m": 6.5,
   "image_input_price_per_1m": 0.45,
   "video_input_price_per_1m": 12.0,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta",
   "input_modalities": [
    "text",
    "image",
    "audio",
    "video"
   ],
   "output_modalities": [
    "embedding"
   ],
   "tags": [
    "embeddings",
    "multimodal",
    "retrieval"
   ],
   "aliases": [
    "gemini-embedding-2-preview",
    "gemini-embedding-2",
    "Gemini Embedding 2"
   ],
   "supports": [
    "batch_api"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini API Models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-08-24",
   "category": "embedding"
  },
  {
   "api_name": "gemini-robotics-er-2-preview",
   "model_name": "Gemini Robotics ER 2",
   "model_family": "Gemini",
   "model_version": "2",
   "provider": "Google",
   "description": "Google's second-generation embodied-reasoning (ER) model for robotics — advanced spatial reasoning, multi-step tool orchestration, agentic code execution, and multi-robot coordination. Reference for Gemini robotics pricing, capability matrix, and migration off ER 1.6 (shut down 2026-08-31).",
   "release_date": "2026-07-30",
   "status": "preview",
   "context_window": 131072,
   "max_output_tokens": 65536,
   "input_price_per_1m": 2.0,
   "cached_input_price_per_1m": 0.2,
   "output_price_per_1m": 10.0,
   "batch_input_price_per_1m": 1.0,
   "batch_output_price_per_1m": 5.0,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/interactions",
   "input_modalities": [
    "text",
    "image",
    "video",
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "robotics",
    "embodied-ai",
    "multimodal",
    "spatial-reasoning",
    "agentic"
   ],
   "aliases": [
    "gemini-robotics-er-2",
    "Gemini Robotics ER 2"
   ],
   "supports": [
    "audio",
    "batch_api",
    "caching",
    "code_execution",
    "computer_use",
    "file_search",
    "function_calling",
    "google_maps_grounding",
    "search_grounding",
    "streaming",
    "structured_outputs",
    "thinking",
    "url_context",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Gemini Robotics ER 2 Preview",
     "url": "https://ai.google.dev/gemini-api/docs/models/gemini-robotics-er-2-preview"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    },
    {
     "label": "Release notes — 2026-07-30",
     "url": "https://ai.google.dev/gemini-api/docs/changelog"
    }
   ],
   "note_modified": "2026-08-10",
   "category": "agent"
  },
  {
   "api_name": "gpt-5.6-cyber",
   "model_name": "GPT-5.6 Cyber",
   "model_family": "GPT",
   "model_version": "5.6",
   "provider": "OpenAI",
   "description": "OpenAI's gated cybersecurity model for approved defenders doing authorized vulnerability research, exploit validation, and security testing. 400K context, Responses API only, requires Daybreak program approval. Reference when scoping defensive security tooling or comparing gated-access OpenAI models.",
   "release_date": "2026-07",
   "knowledge_cutoff": "2026-02",
   "status": "active",
   "context_window": 400000,
   "max_input_tokens": 272000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 12.5,
   "cached_input_price_per_1m": 1.25,
   "output_price_per_1m": 75.0,
   "long_context_threshold": 272000,
   "api_endpoint": "https://api.openai.com/v1/responses",
   "requires_approval": true,
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "reasoning",
    "cybersecurity",
    "gated-access",
    "agentic"
   ],
   "aliases": [
    "daybreak-red-latest",
    "GPT-5.6 Cyber"
   ],
   "supports": [
    "function_calling",
    "prompt_caching",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "gpt-5.6-cyber | OpenAI API Docs",
     "url": "https://developers.openai.com/api/docs/models/gpt-5.6-cyber"
    },
    {
     "label": "OpenAI Models Catalog",
     "url": "https://developers.openai.com/api/docs/models"
    }
   ],
   "note_modified": "2026-08-13",
   "category": "text"
  },
  {
   "api_name": "gpt-6-astra",
   "model_name": "GPT-6 Astra",
   "model_family": "GPT",
   "model_version": "6",
   "provider": "OpenAI",
   "description": "OpenAI's most capable model, opening the GPT-6 line above the GPT-5.6 Sol/Terra/Luna family. Built for the hardest end-to-end work, with a max reasoning effort tier and the broadest hosted-tool set (hosted_shell, apply_patch, skills, computer_use, tool_search). Initially limited to Trusted Access Program enterprises. Reference for OpenAI frontier capability, pricing, and effort levels.",
   "release_date": "2026-09",
   "knowledge_cutoff": "2026-04-30",
   "status": "active",
   "context_window": 1050000,
   "max_input_tokens": 922000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 10.0,
   "cached_input_price_per_1m": 1.0,
   "output_price_per_1m": 50.0,
   "long_context_threshold": 272000,
   "reasoning_effort_levels": [
    "low",
    "medium",
    "high",
    "xhigh",
    "max"
   ],
   "api_endpoint": "https://api.openai.com/v1/responses",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "agentic",
    "flagship"
   ],
   "aliases": [
    "gpt-6-astra",
    "GPT-6 Astra",
    "GPT-6"
   ],
   "supports": [
    "batch_api",
    "function_calling",
    "prompt_caching",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "GPT-6 Astra model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-6-astra"
    }
   ],
   "note_modified": "2026-09-30",
   "category": "text"
  },
  {
   "api_name": "gpt-6-luna",
   "model_name": "GPT-6 Luna",
   "model_family": "GPT",
   "model_version": "6",
   "provider": "OpenAI",
   "description": "OpenAI's most efficient GPT-6 model (released 2026-09-22) for focused, high-volume tasks. 1.05M context, 128K output, none-to-max reasoning effort, $0.10/$0.50 pricing and the latest OpenAI knowledge cutoff (2026-05-18). Reference for low-cost, high-throughput OpenAI workloads.",
   "release_date": "2026-09-22",
   "knowledge_cutoff": "2026-05-18",
   "status": "active",
   "context_window": 1050000,
   "max_input_tokens": 922000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 0.1,
   "cached_input_price_per_1m": 0.01,
   "output_price_per_1m": 0.5,
   "long_context_threshold": 272000,
   "reasoning_effort_levels": [
    "none",
    "low",
    "medium",
    "high",
    "xhigh",
    "max"
   ],
   "default_reasoning_effort": "medium",
   "api_endpoint": "https://api.openai.com/v1/responses",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "efficient",
    "high-volume"
   ],
   "aliases": [
    "gpt-6-luna",
    "GPT-6 Luna"
   ],
   "supports": [
    "batch_api",
    "function_calling",
    "prompt_caching",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "GPT-6 Luna model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-6-luna"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    },
    {
     "label": "TechCrunch launch coverage",
     "url": "https://techcrunch.com/2026/09/22/openai-launches-gpt-6-sol-and-luna/"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "text"
  },
  {
   "api_name": "gpt-6.1-sol",
   "model_name": "GPT-6.1 Sol",
   "model_family": "GPT",
   "model_version": "6.1",
   "provider": "OpenAI",
   "description": "OpenAI's GPT-6 mid tier upgrade (released 2026-09-29 at DevDay), replacing GPT-6 Sol with near-Astra performance on agentic coding, computer use and professional work at one fifth of Astra's price. 1.05M context, 128K output, low-to-max reasoning (no none), $2/$10 with $0.10 cached input. Reference for OpenAI's recommended cost-performance agentic model.",
   "release_date": "2026-09-29",
   "knowledge_cutoff": "2026-04-30",
   "status": "active",
   "context_window": 1050000,
   "max_input_tokens": 922000,
   "max_output_tokens": 128000,
   "input_price_per_1m": 2.0,
   "cached_input_price_per_1m": 0.1,
   "output_price_per_1m": 10.0,
   "long_context_threshold": 272000,
   "reasoning_effort_levels": [
    "low",
    "medium",
    "high",
    "xhigh",
    "max"
   ],
   "default_reasoning_effort": "medium",
   "api_endpoint": "https://api.openai.com/v1/responses",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "coding",
    "agentic"
   ],
   "aliases": [
    "gpt-6.1-sol",
    "GPT-6.1 Sol"
   ],
   "supports": [
    "batch_api",
    "function_calling",
    "prompt_caching",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "GPT-6.1 Sol model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-6.1-sol"
    },
    {
     "label": "Introducing GPT-6.1 Sol",
     "url": "https://openai.com/index/introducing-gpt-6-1-sol/"
    },
    {
     "label": "TechCrunch launch coverage",
     "url": "https://techcrunch.com/2026/09/29/openai-launches-gpt-6-1-sol-says-it-nearly-matches-gpt-6-astra-and-costs-less/"
    }
   ],
   "note_modified": "2026-10-06",
   "category": "text"
  },
  {
   "api_name": "gpt-audio-1.5",
   "model_name": "GPT Audio 1.5",
   "model_family": "GPT Audio",
   "model_version": "1.5",
   "provider": "OpenAI",
   "description": "OpenAI's first GA audio model for audio in/out via Chat Completions API. 128K context with text and audio modalities.",
   "release_date": "2026-02-01",
   "knowledge_cutoff": "2024-09-30",
   "status": "active",
   "context_window": 128000,
   "max_output_tokens": 16384,
   "text_input_price_per_1m": 2.5,
   "text_output_price_per_1m": 10.0,
   "audio_input_price_per_1m": 32.0,
   "audio_output_price_per_1m": 64.0,
   "api_endpoint": "https://api.openai.com/v1/chat/completions",
   "input_modalities": [
    "text",
    "audio"
   ],
   "output_modalities": [
    "text",
    "audio"
   ],
   "tags": [
    "audio",
    "voice",
    "chat",
    "multimodal"
   ],
   "aliases": [
    "gpt-audio-1.5-latest"
   ],
   "supports": [
    "audio",
    "function_calling",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "GPT Audio 1.5",
     "url": "https://developers.openai.com/api/docs/models/gpt-audio-1.5"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-03-07",
   "category": "realtime"
  },
  {
   "api_name": "gpt-image-2.5-flare",
   "model_name": "GPT-Image-2.5 Flare",
   "model_family": "GPT Image",
   "model_version": "2.5",
   "provider": "OpenAI",
   "description": "OpenAI's fast, good-quality everyday image generation model (snapshot 2026-09-08), sibling of GPT-Image-2.5 Sunburst with the same token rates. Reference for lower-latency OpenAI image generation.",
   "release_date": "2026-09-08",
   "status": "active",
   "text_input_price_per_1m": 5.0,
   "text_cached_input_price_per_1m": 1.25,
   "image_input_price_per_1m": 8.0,
   "image_output_price_per_1m": 30.0,
   "api_endpoint": "https://api.openai.com/v1/images/generations",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "image"
   ],
   "tags": [
    "image-generation",
    "image-editing"
   ],
   "aliases": [
    "gpt-image-2.5-flare",
    "gpt-image-2.5-flare-2026-09-08",
    "GPT-Image-2.5 Flare"
   ],
   "quality_levels": [
    "low",
    "medium",
    "high",
    "xhigh",
    "max",
    "auto"
   ],
   "supports": [
    "inpainting",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "GPT-Image-2.5 Flare model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-image-2.5-flare"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "image"
  },
  {
   "api_name": "gpt-image-2.5-sunburst",
   "model_name": "GPT-Image-2.5 Sunburst",
   "model_family": "GPT Image",
   "model_version": "2.5",
   "provider": "OpenAI",
   "description": "OpenAI's most capable image generation and editing model (snapshot 2026-09-08), aimed at precise edits with low-to-max quality tiers. Same token rates as GPT Image 2 ($8 image input, $30 image output per 1M). Reference for top-quality OpenAI image generation.",
   "release_date": "2026-09-08",
   "status": "active",
   "text_input_price_per_1m": 5.0,
   "text_cached_input_price_per_1m": 1.25,
   "image_input_price_per_1m": 8.0,
   "image_output_price_per_1m": 30.0,
   "api_endpoint": "https://api.openai.com/v1/images/generations",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "image"
   ],
   "tags": [
    "image-generation",
    "image-editing"
   ],
   "aliases": [
    "gpt-image-2.5-sunburst",
    "gpt-image-2.5-sunburst-2026-09-08",
    "GPT-Image-2.5 Sunburst"
   ],
   "quality_levels": [
    "low",
    "medium",
    "high",
    "xhigh",
    "max",
    "auto"
   ],
   "supports": [
    "inpainting",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "GPT-Image-2.5 Sunburst model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-image-2.5-sunburst"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-24",
   "category": "image"
  },
  {
   "api_name": "gpt-live-1",
   "model_name": "GPT-Live 1",
   "model_family": "GPT Live",
   "model_version": "1",
   "provider": "OpenAI",
   "description": "OpenAI's premier full-duplex voice model for natural, expressive conversations with smooth interruption handling; it can delegate reasoning and tool calls to a backend agent. Served only via the Live API at $0.05 per session minute. Reference for OpenAI voice agents and ChatGPT Voice (launched 2026-07).",
   "release_date": "2026-07",
   "knowledge_cutoff": "2025-07-31",
   "status": "active",
   "price_per_minute": 0.05,
   "api_endpoint": "https://api.openai.com/v1/live/sessions",
   "input_modalities": [
    "audio",
    "text"
   ],
   "output_modalities": [
    "audio",
    "text"
   ],
   "tags": [
    "voice",
    "realtime",
    "speech-to-speech",
    "audio"
   ],
   "aliases": [
    "gpt-live-1",
    "GPT-Live 1"
   ],
   "supports": [
    "audio",
    "function_calling",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "GPT-Live 1 model page",
     "url": "https://developers.openai.com/api/docs/models/gpt-live-1"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    },
    {
     "label": "How-To Geek: ChatGPT voice on GPT-Live",
     "url": "https://www.howtogeek.com/chatgpt-voice-openai-gpt-live-model/"
    }
   ],
   "note_modified": "2026-09-24",
   "category": "realtime"
  },
  {
   "api_name": "gpt-live-transcribe",
   "model_name": "GPT Live Transcribe",
   "model_family": "GPT",
   "provider": "OpenAI",
   "description": "OpenAI's live/streaming speech-to-text model for realtime transcription ($0.017/min). Listed in both the transcription and realtime/audio catalogs. Reference for streaming STT via the OpenAI Realtime/Audio APIs.",
   "release_date": "2026",
   "status": "active",
   "audio_input_price_per_minute": 0.017,
   "api_endpoint": "https://api.openai.com/v1/realtime",
   "input_modalities": [
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "speech-to-text",
    "transcription",
    "audio",
    "realtime"
   ],
   "aliases": [
    "gpt-live-transcribe",
    "GPT Live Transcribe"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-07-30",
   "category": "transcribe"
  },
  {
   "api_name": "gpt-oss-120b",
   "model_name": "GPT OSS 120B",
   "model_family": "GPT OSS",
   "model_version": "120b",
   "provider": "OpenAI",
   "description": "Open-weight flagship model with 120B parameters",
   "release_date": "2025-08-01",
   "knowledge_cutoff": "2024-06-01",
   "status": "active",
   "context_window": 131072,
   "max_output_tokens": 131072,
   "input_price_per_1m": 0.15,
   "output_price_per_1m": 0.6,
   "license": "Apache 2.0",
   "total_parameters": "117B",
   "active_parameters_per_token": "5.1B",
   "api_endpoint": "https://api.openai.com/v1/chat/completions",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "open-weight",
    "open-source"
   ],
   "aliases": [
    "gpt-oss-120b",
    "gptoss120b"
   ],
   "supports": [
    "function_calling",
    "reasoning",
    "streaming",
    "structured_outputs"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI models",
     "url": "https://developers.openai.com/api/docs/models"
    }
   ],
   "note_modified": "2026-01-14",
   "category": "text"
  },
  {
   "api_name": "gpt-oss-20b",
   "model_name": "GPT OSS 20B",
   "model_family": "GPT OSS",
   "model_version": "20b",
   "provider": "OpenAI",
   "description": "Open-weight efficient model with 20B parameters",
   "release_date": "2025-08-01",
   "knowledge_cutoff": "2024-06-01",
   "status": "active",
   "context_window": 131072,
   "max_output_tokens": 131072,
   "input_price_per_1m": 0.05,
   "output_price_per_1m": 0.2,
   "license": "Apache 2.0",
   "total_parameters": "21B",
   "active_parameters_per_token": "3.6B",
   "num_experts": 32,
   "api_endpoint": "https://api.openai.com/v1/chat/completions",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "open-weight",
    "open-source",
    "efficient"
   ],
   "aliases": [
    "gpt-oss-20b",
    "gptoss20b"
   ],
   "supports": [
    "function_calling",
    "reasoning",
    "streaming",
    "structured_outputs"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI models",
     "url": "https://developers.openai.com/api/docs/models"
    }
   ],
   "note_modified": "2026-01-14",
   "category": "text"
  },
  {
   "api_name": "gpt-realtime-2.1-mini",
   "model_name": "GPT Realtime 2.1 mini",
   "model_family": "GPT Realtime",
   "model_version": "2.1",
   "provider": "OpenAI",
   "description": "OpenAI's cost-efficient realtime voice model (2026) — the mini variant of GPT-Realtime-2.1 for speech-to-speech voice agents at roughly a third of the audio price. Reference for low-cost realtime voice pricing.",
   "knowledge_cutoff": "2024-09",
   "status": "active",
   "context_window": 128000,
   "max_output_tokens": 32000,
   "text_input_price_per_1m": 0.6,
   "text_output_price_per_1m": 2.4,
   "text_cached_input_price_per_1m": 0.06,
   "audio_input_price_per_1m": 10.0,
   "audio_output_price_per_1m": 20.0,
   "audio_cached_input_price_per_1m": 0.3,
   "api_endpoint": "https://api.openai.com/v1/realtime",
   "input_modalities": [
    "text",
    "audio",
    "image"
   ],
   "output_modalities": [
    "text",
    "audio"
   ],
   "tags": [
    "realtime",
    "voice",
    "audio",
    "cost-efficient",
    "multimodal"
   ],
   "aliases": [
    "gpt-realtime-2.1-mini"
   ],
   "supports": [
    "audio",
    "function_calling",
    "reasoning",
    "streaming",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "OpenAI API Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-07-12",
   "category": "realtime"
  },
  {
   "api_name": "gpt-realtime-2.1",
   "model_name": "GPT Realtime 2.1",
   "model_family": "GPT Realtime",
   "model_version": "2.1",
   "provider": "OpenAI",
   "description": "OpenAI's updated GPT-5-class realtime voice model (2026) — refines GPT-Realtime-2 with better alphanumeric recognition, silence/noise handling, and interruption behavior for speech-to-speech voice agents. Reference for the current flagship realtime voice model and its audio pricing.",
   "knowledge_cutoff": "2024-09",
   "status": "active",
   "context_window": 128000,
   "max_output_tokens": 32000,
   "text_input_price_per_1m": 4.0,
   "text_output_price_per_1m": 24.0,
   "text_cached_input_price_per_1m": 0.4,
   "audio_input_price_per_1m": 32.0,
   "audio_output_price_per_1m": 64.0,
   "audio_cached_input_price_per_1m": 0.4,
   "image_input_price_per_1m": 5.0,
   "api_endpoint": "https://api.openai.com/v1/realtime",
   "input_modalities": [
    "text",
    "audio",
    "image"
   ],
   "output_modalities": [
    "text",
    "audio"
   ],
   "tags": [
    "realtime",
    "voice",
    "audio",
    "reasoning",
    "multimodal"
   ],
   "aliases": [
    "gpt-realtime-2.1",
    "gpt-realtime-2.1-latest"
   ],
   "supports": [
    "audio",
    "function_calling",
    "reasoning",
    "streaming",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "gpt-realtime-2.1 Model | OpenAI API",
     "url": "https://developers.openai.com/api/docs/models/gpt-realtime-2.1"
    },
    {
     "label": "OpenAI API Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-24",
   "category": "realtime"
  },
  {
   "api_name": "gpt-realtime-translate",
   "model_name": "GPT Realtime Translate",
   "model_family": "GPT Realtime",
   "model_version": "1",
   "provider": "OpenAI",
   "description": "OpenAI's streaming speech-to-speech translation model for live multilingual audio — translates 70+ input languages into 13 output languages while keeping pace with the speaker. Reference for realtime live-translation use cases and per-minute pricing.",
   "release_date": "2026-05-07",
   "knowledge_cutoff": "2024-09-30",
   "status": "active",
   "context_window": 16000,
   "max_output_tokens": 2000,
   "price_per_minute": 0.034,
   "api_endpoint": "https://api.openai.com/v1/realtime",
   "input_modalities": [
    "audio"
   ],
   "output_modalities": [
    "audio",
    "text"
   ],
   "tags": [
    "realtime",
    "voice",
    "audio",
    "translation"
   ],
   "aliases": [
    "gpt-realtime-translate-latest"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Advancing voice intelligence with new models in the API",
     "url": "https://openai.com/index/advancing-voice-intelligence-with-new-models-in-the-api/"
    },
    {
     "label": "gpt-realtime-translate Model | OpenAI API",
     "url": "https://developers.openai.com/api/docs/models/gpt-realtime-translate"
    },
    {
     "label": "OpenAI API Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-05-30",
   "category": "realtime"
  },
  {
   "api_name": "gpt-rosalind-research",
   "model_name": "GPT-Rosalind",
   "model_family": "GPT",
   "model_version": "1",
   "provider": "OpenAI",
   "description": "OpenAI's life-sciences reasoning model, available only to approved research organizations through a trusted-access program. Priced at $5 input / $25 output per 1M with billing starting 2026-10-05. Reference when evaluating OpenAI models for biology or life-sciences research.",
   "release_date": "2026-09",
   "status": "preview",
   "input_price_per_1m": 5.0,
   "cached_input_price_per_1m": 0.5,
   "output_price_per_1m": 25.0,
   "api_endpoint": "https://api.openai.com/v1/responses",
   "input_modalities": [
    "text"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "reasoning",
    "life-sciences",
    "trusted-access"
   ],
   "aliases": [
    "gpt-rosalind-research",
    "GPT-Rosalind"
   ],
   "supports": [
    "reasoning"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "agent"
  },
  {
   "api_name": "gpt-transcribe",
   "model_name": "GPT Transcribe",
   "model_family": "GPT",
   "provider": "OpenAI",
   "description": "OpenAI's current-generation speech-to-text model, succeeding gpt-4o-transcribe. Simple per-minute pricing ($0.0045/min). Reference for batch audio transcription via the OpenAI Audio API.",
   "release_date": "2026",
   "status": "active",
   "audio_input_price_per_minute": 0.0045,
   "api_endpoint": "https://api.openai.com/v1/audio/transcriptions",
   "input_modalities": [
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "speech-to-text",
    "transcription",
    "audio"
   ],
   "aliases": [
    "gpt-transcribe",
    "GPT Transcribe"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "OpenAI Models",
     "url": "https://developers.openai.com/api/docs/models"
    },
    {
     "label": "OpenAI Pricing",
     "url": "https://developers.openai.com/api/docs/pricing"
    }
   ],
   "note_modified": "2026-07-30",
   "category": "transcribe"
  },
  {
   "api_name": "grok-4.7",
   "model_name": "Grok 4.7",
   "model_family": "Grok",
   "model_version": "4.7",
   "provider": "xAI",
   "description": "SpaceXAI's top-tier Grok model (released 2026-09-21, supersedes Grok 4.6) for coding, agentic tasks, and knowledge work. Larger new base model with a longer RL run, 500K context, no output limit, low-to-xhigh reasoning, and unchanged $2/$6 pricing. Reference for the current best xAI API model.",
   "release_date": "2026-09-21",
   "knowledge_cutoff": "2026-05",
   "status": "active",
   "context_window": 500000,
   "input_price_per_1m": 2.0,
   "cached_input_price_per_1m": 0.5,
   "output_price_per_1m": 6.0,
   "input_price_per_1m_long": 4.0,
   "output_price_per_1m_long": 12.0,
   "long_context_threshold": 200000,
   "reasoning_effort_levels": [
    "low",
    "medium",
    "high",
    "xhigh"
   ],
   "api_endpoint": "https://api.x.ai/v1/responses",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "chat",
    "text-generation",
    "reasoning",
    "flagship",
    "coding",
    "agentic"
   ],
   "aliases": [
    "grok-4-7",
    "grok-4.7-latest",
    "Grok 4.7"
   ],
   "supports": [
    "function_calling",
    "reasoning",
    "streaming",
    "structured_outputs",
    "vision"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "xAI Models",
     "url": "https://docs.x.ai/docs/models"
    },
    {
     "label": "Grok 4.7 docs",
     "url": "https://docs.x.ai/developers/grok-4-7"
    },
    {
     "label": "MarkTechPost launch coverage",
     "url": "https://www.marktechpost.com/2026/09/21/spacexai-releases-grok-4-7/"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "text"
  },
  {
   "api_name": "grok-imagine-image-2.0",
   "model_name": "Grok Imagine Image 2.0",
   "model_family": "Grok Imagine",
   "model_version": "2.0",
   "provider": "xAI",
   "description": "xAI's second-generation Grok Imagine image model at $0.04/image, sitting between the $0.02 base model and the $0.05 quality tier. Sparsely documented — only pricing is published. Reference when choosing an xAI image model by cost tier.",
   "release_date": "2026-08",
   "status": "active",
   "price_per_image": 0.04,
   "api_endpoint": "https://api.x.ai/v1/images/generations",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "image"
   ],
   "tags": [
    "image-generation",
    "creative"
   ],
   "aliases": [
    "Grok Imagine Image 2.0"
   ],
   "supports": [],
   "is_sota": true,
   "sources": [
    {
     "label": "xAI Models and Pricing",
     "url": "https://docs.x.ai/docs/models"
    }
   ],
   "note_modified": "2026-09-23",
   "category": "image"
  },
  {
   "api_name": "grok-imagine-video-1.5",
   "model_name": "Grok Imagine Video 1.5",
   "model_family": "Grok Imagine",
   "model_version": "1.5",
   "provider": "xAI",
   "description": "xAI's upgraded Grok Imagine video generation model (1.5), listed on the official docs.x.ai lineup at $0.080/second. Reference for current xAI video generation pricing and capabilities.",
   "release_date": "2026-05-01",
   "status": "active",
   "price_per_minute": 4.8,
   "price_per_second": 0.08,
   "api_endpoint": "https://api.x.ai/v1/videos/generations",
   "input_modalities": [
    "text",
    "image",
    "video"
   ],
   "output_modalities": [
    "video",
    "audio"
   ],
   "tags": [
    "video-generation",
    "creative",
    "multimodal"
   ],
   "aliases": [
    "Grok Imagine Video 1.5"
   ],
   "supports": [
    "audio"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "xAI Models",
     "url": "https://docs.x.ai/docs/models"
    }
   ],
   "note_modified": "2026-06-20",
   "category": "video"
  },
  {
   "api_name": "grok-voice-think-fast-2.0",
   "model_name": "Grok Voice Think Fast 2.0",
   "model_family": "Grok",
   "model_version": "2.0",
   "provider": "xAI",
   "description": "xAI's newest speech-to-speech voice model (fast-thinking tier, v2.0). Billed per minute of audio ($0.08/min) plus text input ($0.004). Reference for Grok voice agent integrations at the top voice tier.",
   "status": "active",
   "api_endpoint": "https://api.x.ai/v1",
   "input_modalities": [
    "audio",
    "text"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "speech-to-speech",
    "voice",
    "audio"
   ],
   "aliases": [
    "grok-voice-think-fast-2.0",
    "Grok Voice Think Fast 2.0"
   ],
   "supports": [
    "audio",
    "streaming"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "xAI Models",
     "url": "https://docs.x.ai/docs/models"
    }
   ],
   "note_modified": "2026-07-30",
   "category": "realtime"
  },
  {
   "api_name": "lyria-3.5",
   "model_name": "Lyria 3.5",
   "model_family": "Lyria",
   "model_version": "3.5",
   "provider": "Google",
   "description": "Google's flagship music generation model (stable), optimized for full-length songs with verses, choruses, and bridges. Generates 44.1 kHz stereo MP3 or WAV from text prompts plus up to 10 reference images, via the Interactions API. Billed per song at $0.08. Reference for AI music generation length limits, formats, and Lyria variant selection.",
   "release_date": "2026-07-29",
   "status": "active",
   "price_per_song": 0.08,
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta/interactions",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "audio"
   ],
   "tags": [
    "music-generation",
    "audio",
    "generative-media"
   ],
   "aliases": [
    "lyria-3.5",
    "Lyria 3.5"
   ],
   "output_formats": [
    "mp3",
    "wav"
   ],
   "supports": [
    "watermark"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Generate music with Lyria 3.5",
     "url": "https://ai.google.dev/gemini-api/docs/music-generation"
    },
    {
     "label": "Gemini API Pricing",
     "url": "https://ai.google.dev/gemini-api/docs/pricing"
    }
   ],
   "note_modified": "2026-09-04",
   "category": "music"
  },
  {
   "api_name": "scribe_v2",
   "model_name": "Scribe v2",
   "model_family": "Scribe",
   "model_version": "v2",
   "provider": "ElevenLabs",
   "description": "ElevenLabs flagship Speech-to-Text model with >95% WER accuracy across 90+ languages. Supports word-level timestamps, smart speaker diarization (up to 48 speakers), entity detection (56 types), and keyterm prompting. Reference for batch transcription workloads — meetings, podcasts, audiobooks.",
   "release_date": "2025-09-01",
   "status": "active",
   "context_window": 0,
   "max_output_tokens": 0,
   "price_per_hour": 0.22,
   "api_endpoint": "https://api.elevenlabs.io/v1/speech-to-text",
   "input_modalities": [
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "stt",
    "transcription",
    "audio",
    "speech-to-text",
    "elevenlabs",
    "diarization"
   ],
   "aliases": [
    "scribe-v2",
    "elevenlabs-scribe-v2"
   ],
   "supports": [
    "audio",
    "diarization",
    "language_detection",
    "structured_outputs",
    "word_timestamps"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "ElevenLabs models",
     "url": "https://elevenlabs.io/docs/overview/models"
    }
   ],
   "note_modified": "2026-05-14",
   "category": "transcribe"
  },
  {
   "api_name": "scribe_v2_realtime",
   "model_name": "Scribe v2 Realtime",
   "model_family": "Scribe",
   "model_version": "v2-realtime",
   "provider": "ElevenLabs",
   "description": "ElevenLabs realtime Speech-to-Text model with ~150ms latency across 90+ languages. Supports streaming, voice activity detection, and automatic language recognition. Reference when building voice agents, live captioning, or any conversational STT use case.",
   "release_date": "2025-09-01",
   "status": "active",
   "context_window": 0,
   "max_output_tokens": 0,
   "price_per_hour": 0.39,
   "api_endpoint": "wss://api.elevenlabs.io/v1/speech-to-text/stream",
   "input_modalities": [
    "audio"
   ],
   "output_modalities": [
    "text"
   ],
   "tags": [
    "stt",
    "transcription",
    "audio",
    "speech-to-text",
    "elevenlabs",
    "realtime",
    "streaming"
   ],
   "aliases": [
    "scribe-v2-realtime",
    "elevenlabs-scribe-realtime"
   ],
   "supports": [
    "audio",
    "language_detection",
    "streaming",
    "word_timestamps"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "ElevenLabs models",
     "url": "https://elevenlabs.io/docs/overview/models"
    }
   ],
   "note_modified": "2026-07-13",
   "category": "transcribe"
  },
  {
   "api_name": "veo-3.1-generate-preview",
   "model_name": "Veo 3.1",
   "model_family": "Veo",
   "model_version": "3.1",
   "provider": "Google",
   "description": "Google's state-of-the-art cinematic video generation engine for professional-grade 4K output, natively synchronized audio, and complex camera movements",
   "release_date": "2026-01-01",
   "status": "preview",
   "api_endpoint": "https://generativelanguage.googleapis.com/v1beta",
   "input_modalities": [
    "text",
    "image"
   ],
   "output_modalities": [
    "video"
   ],
   "tags": [
    "video-generation",
    "cinematic",
    "4k",
    "audio-sync",
    "creative"
   ],
   "aliases": [
    "veo-3.1-fast-generate-preview"
   ],
   "supports": [
    "audio_generation",
    "video"
   ],
   "is_sota": true,
   "sources": [
    {
     "label": "Veo 3.1 Documentation",
     "url": "https://ai.google.dev/gemini-api/docs/models/veo-3.1-generate-preview"
    },
    {
     "label": "Gemini API Models",
     "url": "https://ai.google.dev/gemini-api/docs/models"
    }
   ],
   "note_modified": "2026-03-07",
   "category": "video"
  }
 ]
}