• English

      • Providers

      Savfox supports 60+ cloud and local model providers via OpenAI-compatible API endpoints.

      Provider Guides

      API key/OAuth auth and model selection. Claude configuration and provider-specific headers. Gemini models via Google AI or Vertex AI. DeepSeek models for coding and reasoning. Ultra-fast inference with Groq hardware. Local models via OpenAI-compatible endpoint.

      Selection Tips

      • Coding and general usage: gpt-4.1, claude-sonnet-4, deepseek-coder
      • Fast low-cost runs: gpt-4.1-mini, claude-haiku, Groq models
      • Complex reasoning: o3, claude-opus-4, deepseek-reasoner
      • Local-first/privacy: Ollama, LM Studio

      Other Supported Providers

      Savfox works with any OpenAI-compatible API endpoint. Popular providers include:

      ProviderProvider IDNotes
      xAI (Grok)xaihttps://api.x.ai/v1
      Mistralmistralhttps://api.mistral.ai/v1
      OpenRouteropenrouterAggregator with 100+ models
      Together AItogetheraiOpen-source model hosting
      Fireworks AIfireworks-aiFast inference
      CerebrascerebrasUltra-fast inference
      PerplexityperplexitySearch-augmented models
      SiliconFlowsiliconflowChinese provider
      Moonshot AImoonshotaiKimi models
      VolcenginevolcengineByteDance cloud
      ZhipuAIzhipuaiGLM models

      Configure any provider by setting provider_id and optionally base_url in your config:

      [model]
provider_id = "xai"
model = "grok-3"
api_key = "${XAI_API_KEY}"