Skip to content

Available Model List ​

This page lists current mainstream AI provider models and capabilities for lookup and integration. For pricing, see Model Pricing.

Model Notes ​

  • The system uses the latest available models from each AI provider by default
  • Different API keys are mapped to different provider groups
  • Model availability depends on group authorization
  • Model calls stay highly compatible with official APIs

Available Model List ​

OpenAI (GPT Series) ​

Focuses on deep reasoning and native multimodal capabilities, evolving towards strong logic and complex system coordination.

Model NameMax Input / Output TokensOperating ModeOpenAI Protocol
gpt-6.2-sol922K / 128Kchat✅ Supported
gpt-6.2-astra922K / 128Kchat✅ Supported
gpt-6.1-sol922K / 128Kchat✅ Supported
gpt-6.1-terra922K / 128Kchat✅ Supported
gpt-6-sol922K / 128Kchat✅ Supported
gpt-6-luna922K / 128Kchat✅ Supported
gpt-6-astra922K / 128Kchat✅ Supported
gpt-5.6-sol1050K / 128Kchat✅ Supported
gpt-5.6-terra1050K / 128Kchat✅ Supported
gpt-5.6-luna1050K / 128Kchat✅ Supported
gpt-5.51050K / 128Kchat✅ Supported
gpt-image-2.5-flare- / -image_generation✅ Supported
gpt-image-2.5-sunburst- / -image_generation✅ Supported
gpt-image-2-medium- / -image_generation✅ Supported
gpt-image-1.5- / -image_generation✅ Supported
gpt-image-1- / -image_generation✅ Supported

Anthropic (Claude Series) ​

Focuses on long-text fidelity, elegant code architecture, and highly human-like natural writing style.

Model NameMax Input / Output TokensOperating ModeOpenAI Protocol
claude-fable-5-21000K / 128Kchat✅ Supported
claude-fable-5-11000K / 128Kchat✅ Supported
claude-fable-51000K / 128Kchat✅ Supported
claude-opus-5-51000K / 128Kchat✅ Supported
claude-opus-51000K / 128Kchat✅ Supported
claude-opus-4-81000K / 128Kchat✅ Supported
claude-opus-4-71000K / 128Kchat✅ Supported
claude-opus-4-61000K / 128Kchat✅ Supported
claude-opus-4-5200K / 64Kchat✅ Supported
claude-sonnet-5-51000K / 128Kchat✅ Supported
claude-sonnet-51000K / 128Kchat✅ Supported
claude-sonnet-4-61000K / 128Kchat✅ Supported
claude-sonnet-4-5200K / 64Kchat✅ Supported
claude-haiku-5-51000K / 128Kchat✅ Supported
claude-haiku-4-5-20251001200K / 64Kchat✅ Supported

Google (Gemini Series) ​

Focuses on native ultra-large context throughput, deep web search (Deep Research), and Google ecosystem integration.

Model NameMax Input / Output TokensOperating ModeOpenAI Protocol
gemini-4-argon1048K / 1048Kchat✅ Supported
gemini-3.8-flash1048K / 65Kchat✅ Supported
gemini-3.8-flash-lite-tts8K / 16Kaudio_speech⚠️ Native only
gemini-3.8-flash-tts8K / 16Kaudio_speech⚠️ Native only
gemini-3.7-flash1048K / 65Kchat✅ Supported
gemini-3.5-flash1048K / 65Kchat✅ Supported
gemini-3.5-flash-lite1048K / 65Kchat✅ Supported
gemini-3.5-transcribe98K / 32Kaudio_transcription⚠️ Native only
gemini-3.1-flash-lite1048K / 65Kchat✅ Supported
gemini-3.1-pro-preview1048K / 65Kchat✅ Supported
gemini-3.1-pro-preview-customtools1048K / 65Kchat✅ Supported
gemini-3-flash-preview1048K / 65Kchat✅ Supported
gemini-2.5-flash1048K / 65Kchat✅ Supported
gemini-2.5-flash-lite1048K / 65Kchat✅ Supported
gemini-2.5-pro1048K / 65Kchat✅ Supported
gemini-2.5-flash-preview-tts8K / 16Kaudio_speech⚠️ Native only
gemini-nano-banana-2.1131K / 32Kimage_generation⚠️ Native only
gemini-robotics-er-2-preview131K / 65Kchat✅ Supported
gemini-3.1-flash-image131K / 32Kimage_generation⚠️ Native only
gemini-3.1-flash-lite-image65K / 4Kimage_generation⚠️ Native only
gemini-3-pro-image65K / 32Kimage_generation⚠️ Native only
gemini-2.5-flash-image32K / 32Kimage_generation⚠️ Native only

Image, speech and transcription models use the native Gemini endpoint POST /v1beta/models/{model}:generateContent. For speech, set responseModalities: ["AUDIO"]; transcription text is returned in audioTranscription.text.

Some names are channel compatibility IDs: native API checks returned gemini-3.6-flash for gemini-3.5-flash and gemini-3.8-flash for gemini-3.7-flash. Check the response's modelVersion for the actual version.

xAI (Grok Series) ​

Focuses on long-context multimodal reasoning, tool use, and web-assisted workflows for complex Q&A, code analysis, and research tasks.

Model NameMax Input / Output TokensOperating ModeOpenAI Protocol
grok-4.5500K / 500Kchat✅ Supported
grok-4.31000K / 1000Kchat✅ Supported

Notes ​

  • Gemini last tested: 2026-11-18; all 22 listed models were checked for usable output and actual billing.

  • "OpenAI Protocol" column: ✅ indicates OpenAI-compatible text endpoints (such as /v1/chat/completions); ⚠️ Native only means image, speech and transcription models use Google native :generateContent.

  • Gemini image models (5 current IDs): gemini-nano-banana-2.1, gemini-3.1-flash-image, gemini-3.1-flash-lite-image, gemini-3-pro-image, gemini-2.5-flash-image. Use the Google native endpoint to receive images.

  • Full list last checked: 2026-10-28; claude-fable-5-2 added on 2026-11-10. Query the Available Models API with your API key for current availability; one group’s inventory does not represent all groups.

  • The list uses primary model IDs. Some models also expose dated or other compatibility aliases; use the IDs returned by the API. Legacy model availability depends on the selected channel.

  • Model names and capabilities stay aligned with official APIs

  • New models are added when released