The AI models that power your OpenClaw agent's intelligence
OpenClaw model providers are the LLM APIs that supply the intelligence behind your agent. OpenClaw supports multiple providers including Anthropic (Claude), OpenAI (GPT-4), Google (Gemini), and others via OpenRouter. You choose which model powers your agent based on cost, capability, and speed requirements.
Each model provider requires an API key that you configure in OpenClaw's gateway settings. The agent sends context and prompts to the provider's API and receives responses. You can configure fallback models (if primary is unavailable) and model selection rules (use cheap model for simple tasks, powerful model for complex ones).
Model selection has the single biggest impact on both capability and cost. Claude Sonnet is excellent for nuanced tasks. Gemini Flash is fast and cheap for simple operations. Choosing the right model for each task can reduce costs by 10x while maintaining quality. Clawfleet's smart routing feature automatically selects the optimal model per request.
Clawfleet manages your OpenClaw instance — OpenClaw Model Providers, backups, restarts, and cost tracking — all included. Start for $1.
Deploy for $1 →