Model provider quickstart

Model Providers

OpenClaw can use many LLM providers. Pick one, authenticate, then set the default model as

Quick start (two steps)

1. Authenticate with the provider (usually via
   text

   Copy

   openclaw onboard
   ).
2. Set the default model:

json5
Copy
{
  agents: { defaults: { model: { primary: "anthropic/claude-opus-4-6" } } },
}

Supported providers (starter set)

• Alibaba Model Studio
• Amazon Bedrock
• Anthropic (API + Claude CLI)
• BytePlus (International)
• Chutes
• ComfyUI
• Cloudflare AI Gateway
• DeepInfra
• fal
• Fireworks
• GLM models
• MiniMax
• Mistral
• Moonshot AI (Kimi + Kimi Coding)
• OpenAI (API + Codex)
• OpenCode (Zen + Go)
• OpenRouter
• Qianfan
• Qwen
• Runway
• StepFun
• Synthetic
• Vercel AI Gateway
• Venice (Venice AI)
• xAI
• Z.AI

Additional bundled provider variants

• text

  Copy

  anthropic-vertex
  - implicit Anthropic on Google Vertex support when Vertex credentials are available; no separate onboarding auth choice
• text

  Copy

  copilot-proxy
  - local VS Code Copilot Proxy bridge; use
  text

  Copy

  openclaw onboard --auth-choice copilot-proxy
• text

  Copy

  google-gemini-cli
  - unofficial Gemini CLI OAuth flow; requires a local
  text

  Copy

  gemini
  install (
  text

  Copy

  brew install gemini-cli
  or
  text

  Copy

  npm install -g @google/gemini-cli
  ); default model
  text

  Copy

  google-gemini-cli/gemini-3-flash-preview
  ; use
  text

  Copy

  openclaw onboard --auth-choice google-gemini-cli
  or
  text

  Copy

  openclaw models auth login --provider google-gemini-cli --set-default

For the full provider catalog (xAI, Groq, Mistral, etc.) and advanced configuration, see Model providers.

Related

• Model selection
• Model failover
• Models CLI