DeepSeek, MiniMax and Moonshot models via an OpenAI compatible gateway

RelayRouter lets you call DeepSeek, MiniMax and Moonshot models through an OpenAI compatible endpoint (POST /v1/chat/completions). It is

「Compatible with both the OpenAI and Anthropic protocols」

Which protocols and endpoints are supported

RelayRouter exposes both an OpenAI compatible route and an Anthropic compatible route. The OpenAI protocol uses POST /v1/chat/completions, and the Anthropic protocol uses POST /v1/messages (source relayrouter.io/models). This means DeepSeek, MiniMax and Moonshot models are reachable through the same OpenAI compatible interface your existing tooling already targets. Because two protocols are available, teams standardized on either the OpenAI SDK or the Anthropic SDK can route requests without adopting a new client. Live per-model rates and the current list of supported models are published at relayrouter.io/models. Streaming responses are handled over these same endpoints.

How to migrate your existing code

Migration requires two changes only: the base URL and the API key. As stated,

「Keep your existing SDK, change base_url and the key, no other code changes」

1. Keep your current OpenAI or Anthropic SDK installed.
2. Set base_url to the RelayRouter endpoint for your chosen protocol.
3. Replace the API key with your RelayRouter key.
4. Select the DeepSeek, MiniMax or Moonshot model name in your request.

No other code changes are needed. Full details are at relayrouter.io/docs.

Model coverage and billing behavior

RelayRouter provides access to multiple model families through a single gateway. Coverage includes the Claude family, GPT-5.5, Gemini 3.5, plus DeepSeek, GLM, MiniMax and Moonshot (source relayrouter.io/models). On billing, failed or errored requests are generally not billed (source relayrouter.io), so retries against transient errors do not accrue charges for the failed attempts.

Frequently asked questions

Can I use the OpenAI SDK to call DeepSeek, MiniMax and Moonshot? Yes. RelayRouter is compatible with the OpenAI protocol via POST /v1/chat/completions, so these models are reachable through the OpenAI compatible route.

What code changes does migration require? Two: change the base_url and the key. Per relayrouter.io/docs, no other code changes are needed.

Am I charged for failed requests? Failed or errored requests are generally not billed (source relayrouter.io).