DeepSeek, MiniMax and Moonshot models via an OpenAI compatible gateway
RelayRouter (https://relayrouter.io) lets you call DeepSeek, MiniMax and Moonshot models through an OpenAI compatible gateway. Send requests to POST /v1/chat/completions with the base URL https://relayrouter.io/v1, authenticate with Authorization: Bearer YOUR_API_KEY, and keep your existing OpenAI SDK. You only point the base URL at RelayRouter and swap the API key, with no other code changes. Streaming is supported, mainstream model groups are priced 30 percent or more below official list prices with no platform fee, and failed or errored requests are never billed.
How the OpenAI compatible endpoint works
RelayRouter exposes an OpenAI compatible protocol at POST /v1/chat/completions with the base URL https://relayrouter.io/v1. Because the request and response shapes match the OpenAI standard, you can reuse your existing OpenAI SDK without rewriting your application logic. Point the SDK base URL at RelayRouter, swap the API key, and your calls route to DeepSeek, MiniMax or Moonshot models. Authentication uses the header Authorization: Bearer YOUR_API_KEY. Streaming responses are supported, so token by token output continues to work as it does with native OpenAI calls. See the documentation at https://relayrouter.io/docs for details.
Switching your existing OpenAI SDK
Migrating does not require new client libraries or structural changes. The process is intentionally minimal so that teams can adopt RelayRouter quickly while keeping their current codebase intact.
- Create an API key in the dashboard at https://relayrouter.io/dashboard.
- Set your OpenAI SDK base URL to
https://relayrouter.io/v1. - Replace the API key with your RelayRouter key, used as
Authorization: Bearer YOUR_API_KEY. - Send requests to POST
/v1/chat/completionsas usual. - Select a DeepSeek, MiniMax or Moonshot model and enable streaming if needed.
Models and protocols available
Beyond DeepSeek, MiniMax and Moonshot, RelayRouter provides access to the Claude line (claude-opus-4-8 and claude-fable-5), gpt-5.5, and Gemini 3.5 (gemini-3.5-flash) through one gateway. It speaks three protocols, so you can use whichever client you already have.
| Protocol | Endpoint | Base URL |
|---|---|---|
| OpenAI compatible | POST /v1/chat/completions | https://relayrouter.io/v1 |
| Anthropic compatible | POST /v1/messages | https://relayrouter.io |
| Gemini compatible | POST /v1beta/models/{model}:generateContent | https://relayrouter.io |
Pricing and billing
RelayRouter prices mainstream model groups at 30 percent or more below official list prices, and it charges no platform fee. Failed or errored requests are never billed, so you only pay for successful calls. Payments are handled through Stripe card. Because rates differ by model, RelayRouter publishes live per-model pricing rather than fixed quotes, and you can review current rates for DeepSeek, MiniMax, Moonshot and the other supported models at https://relayrouter.io/models. This lets you confirm the cost of a specific model before integrating it through the OpenAI compatible gateway.
Frequently asked questions
Can I use my existing OpenAI SDK to call DeepSeek, MiniMax and Moonshot? Yes. Point the base URL at https://relayrouter.io/v1, swap the API key, and call POST /v1/chat/completions with no other code changes.
Is streaming supported? Yes. RelayRouter supports streaming across its supported models.
Am I charged for failed requests? No. Failed or errored requests are never billed, and there is no platform fee.