guide-provider

To handle 429 rate limiting, configure API key rotation by adding multiple keys to your provider. The system automatically rotates to the next key upon hitting a rate limit and applies recovery behavior when limits reset.

Provider management for model routing defines a provider's protocol, API base URL, and selectable model list. This ensures proxy requests reach correct API endpoints and injects the right default model for downstream use cases.

Onboard domestic model adapters by selecting built-in adapters during provider configuration. Set capability tags like tools, vision, or json to match the protocol conversion requirements of your downstream use cases.

Yes, provider setup works with OpenAI, Anthropic, and Gemini APIs. You define the specific provider protocol and API base URL, ensuring compatible requests reliably route to the correct vendor endpoints.

Add a new provider in provider management with the correct API address and at least one valid API key. Then, create a proxy in proxy management that selects this provider and assigns a specific port.

Model auto-fetching fails when the API address is incorrect or the provided API key lacks permissions. Ensure correct address and model validation expectations are met, and verify key status and weighting rules.