Concepts and configuration
Model providers
Reference for LLM/model providers (not chat channels like WhatsApp/Telegram). For model selection rules, see Models.
This page is an index. The provider reference is documented on four pages, one per reader job. Open the page that matches your task.
| Page | Read it when |
|---|---|
| Quick rules | You need model refs, CLI helpers, or the rules that decide your primary model and OpenAI runtime. |
| Control UI and API keys | You are configuring providers from Settings -> Models, or setting up multiple API keys and rotation. |
| Official provider plugins | You are setting up a bundled provider, or need its id, auth env, example model, and quirks. |
| Custom providers and local runtimes | You are configuring a provider through models.providers, a custom base URL, a proxy, or a local inference server. |
Where each section moved
Every section heading from the previous single-page version keeps its anchor
here, so an existing link such as /concepts/model-providers#byteplus-international still resolves.
Each entry points at the page that now holds the content.
- Quick rules
- Model refs and CLI helpers
- Adding provider auth does not change your primary model
- OpenAI provider/runtime split
- CLI runtimes
- Configure providers in the Control UI
- Plugin-owned provider behavior
- API key rotation
- Key sources and priority
- When rotation kicks in
- Official provider plugins
- OpenAI
- Anthropic
- OpenAI ChatGPT/Codex OAuth
- Other subscription-style hosted options
- OpenCode
- Google Gemini (API key)
- Google Vertex and Gemini CLI runtime
- Z.AI (GLM)
- Vercel AI Gateway
- Other bundled provider plugins
- Quirks worth knowing
- OpenRouter
- Kilo Gateway
- MiniMax (quirks)
- NVIDIA
- xAI
- Providers via
models.providers(custom/base URL) - Moonshot AI (Kimi)
- Kimi Coding
- Volcano Engine (Doubao)
- Standard models (Volcano Engine)
- Coding models (volcengine-plan)
- BytePlus (International)
- Standard models (BytePlus)
- Coding models (byteplus-plan)
- Synthetic
- MiniMax
- llama.cpp
- LM Studio
- Ollama
- vLLM
- SGLang
- Local proxies (LM Studio, vLLM, LiteLLM, etc.)
- Default optional fields
- Proxy-route shaping rules
CLI examples
openclaw onboard --auth-choice opencode-zenopenclaw models set opencode/claude-opus-4-6openclaw models listSee also: Configuration for full configuration examples.
Related
- Configuration reference - model config keys
- Model failover - fallback chains and retry behavior
- Models - model configuration and aliases
- Providers - per-provider setup guides
- Agent harness plugins - SDK surface for plugins that replace the embedded agent executor
openclaw models- list, select, and authenticate providers from the CLI
Was this useful?
