接口文档

OpenAI 兼容 REST API · 流式 SSE · 多模型统一接入

LiteLLM 网关完全兼容 OpenAI REST API。

示例可直接复制运行。

接口地址（Base URL）

https://relay.2frs.com/v1

鉴权

所有接口使用 HTTP Bearer Token 鉴权。

Authorization: Bearer sk-lit-xxxxxxxxxxxxxxxxxxxxxxxx

模型列表

返回当前 Key 有权访问的全部模型。

curl https://relay.2frs.com/v1/models \
  -H "Authorization: Bearer $LITELLM_API_KEY"

{
  "object": "list",
  "data": [
    {"id": "deepseek-flash", "object": "model", "owned_by": "litellm"},
    {"id": "deepseek-v4-pro", "object": "model", "owned_by": "litellm"}
  ]
}

可用模型由白名单与启用渠道共同决定。

对话补全

POST /v1/chat/completions，默认返回完整 JSON。

curl -X POST https://relay.2frs.com/v1/chat/completions \
  -H "Authorization: Bearer $LITELLM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-flash",
    "stream": false,
    "messages": [
      {"role": "user", "content": "用一句话介绍你自己"}
    ]
  }'

非流式响应（节选）

{
  "id": "req_8f2c…",
  "object": "chat.completion",
  "choices": [{
    "index": 0,
    "message": {"role": "assistant", "content": "……"},
    "finish_reason": "stop"
  }],
  "usage": {"prompt_tokens": 12, "completion_tokens": 34, "total_tokens": 46}
}

流式响应

显式 stream:true 时以 SSE 逐块返回。

思考帧与正文帧同一 id，思考过程不计入正文。

curl -N -X POST https://relay.2frs.com/v1/chat/completions \
  -H "Authorization: Bearer $LITELLM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-flash",
    "stream": true,
    "messages": [{"role": "user", "content": "你好"}]
  }'

data: {"choices":[{"delta":{"role":"assistant","content":""}}]}

data: {"choices":[{"delta":{"reasoning_content":"正在思考…"}}]}

data: {"choices":[{"delta":{"content":"你好"}}]}

data: [DONE]

错误码

统一错误结构。

{ "error": { "message": "…", "type": "authentication_error",
  "code": "AUTH002", "request_id": "req_…" } }

计费说明

按真实 usage 结算，倍率体系。

仅按真实用量计费。