Model gateway
The model gateway is a separate service from the Session API. It gives you one API key and one Anthropic-compatible endpoint that can reach DeepSeek, Kimi, and Claude, so a client needs no per-provider account and no provider SDK.
Base: https://api.xiaohe.team/v1 (Anthropic Messages API)
| Method | Path | Purpose |
|---|---|---|
| POST | /v1/messages | Chat completion (Anthropic Messages API) |
| GET | /v1/models | List routed models |
Send the gateway key as x-api-key (or Authorization: Bearer …), pick any routed model, and the gateway forwards the request to the provider configured behind it:
{
"model": "deepseek-v4-pro",
"max_tokens": 1024,
"messages": [{"role": "user", "content": "Hello"}]
}
Routed models include DeepSeek (deepseek-v4-pro, deepseek-chat, deepseek-reasoner), Kimi (kimi-k2.6, kimi-for-coding), and Anthropic (claude-opus-5, claude-sonnet-5, claude-haiku-4-5, claude-fable-5). Usage is recorded per request and billed by token, with costs estimated from the built-in model pricing table.
Adding a provider key of your own to the Agent is a different route to the same models — see Advanced. Both work; the gateway is the one that needs no provider account on the client side.
Related
- Session API — the control plane, a different service with a different key
- Overview — the four credentials
- Advanced — local model providers