Chat Completions 聊天对话

Chat Completions 是 AISP 最核心的接口，支持多轮对话、流式输出（SSE）、Function Calling、多模态（Vision / 图片 / 音频 / 视频输入）等能力。 接口完全兼容 OpenAI Chat Completions API，可无缝替换 OpenAI SDK。

请求头 Headers

请求参数

messages 格式

[
  { "role": "system", "content": "你是一名AI助手" },
  { "role": "user", "content": "你好" },
  { "role": "assistant", "content": "你好，请问有什么可以帮助你？" }
]

Role 说明

普通请求示例

POST /v1/chat/completions
Authorization: Bearer sk-xxxxxxxx
Content-Type: application/json

{
  "model": "deepseek-v4-pro",
  "messages": [
    { "role": "user", "content": "你好，介绍一下你自己" }
  ]
}

普通返回示例

{
  "id": "chatcmpl-xxxx",
  "object": "chat.completion",
  "created": 1710000000,
  "model": "deepseek-v4-pro",
  "choices": [
    {
      "index": 0,
      "message": { "role": "assistant", "content": "你好！我是..." },
      "finish_reason": "stop"
    }
  ]
}

流式输出

将 stream 设为 true，使用 SSE 协议返回：

POST /v1/chat/completions

{
  "model": "deepseek-v4-pro",
  "stream": true,
  "messages": [
    { "role": "user", "content": "你好" }
  ]
}

data: {
  "id": "chatcmpl-xxx",
  "object": "chat.completion.chunk",
  "model": "deepseek-v4-pro",
  "choices": [
    { "index": 0, "delta": { "role": "assistant", "content": "你" }, "finish_reason": null }
  ]
}

data: {
  "id": "chatcmpl-xxx",
  "object": "chat.completion.chunk",
  "model": "deepseek-v4-pro",
  "choices": [
    { "index": 0, "delta": { "content": "好" }, "finish_reason": null }
  ]
}

data: {
  "id": "chatcmpl-xxx",
  "object": "chat.completion.chunk",
  "model": "deepseek-v4-pro",
  "choices": [
    { "index": 0, "delta": {}, "finish_reason": "stop" }
  ]
}

data: [DONE]

Tool Calling（工具调用）

接口支持 OpenAI 标准的 Tool Calling / Function Calling，可用于 Claude Code、Cursor、Trae 等 AI Agent 与编程工具的「执行工具 → 返回结果 → 继续回答」循环。 以下为非流式调用示例（流式调用只需把 stream 设为 true，工具增量会以 delta.tool_calls 分片返回，结束时 finish_reason=tool_calls）。

POST /v1/chat/completions
Authorization: Bearer sk-xxxxxxxx
Content-Type: application/json

{
  "model": "deepseek-v4-flash",
  "tools": [
    {
      "type": "function",
      "function": {
        "name": "get_weather",
        "description": "查询指定城市的天气",
        "parameters": {
          "type": "object",
          "properties": { "city": { "type": "string", "description": "城市名" } },
          "required": ["city"]
        }
      }
    }
  ],
  "messages": [
    { "role": "user", "content": "查询北京的天气" }
  ]
}

模型决定调用工具时，返回 finish_reason=tool_calls 与 message.tool_calls：

{
  "id": "chatcmpl-xxxx",
  "object": "chat.completion",
  "created": 1710000000,
  "model": "deepseek-v4-flash",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": null,
        "tool_calls": [
          {
            "id": "call_xxxx",
            "type": "function",
            "function": { "name": "get_weather", "arguments": "{\"city\": \"北京\"}" }
          }
        ]
      },
      "finish_reason": "tool_calls"
    }
  ]
}

客户端执行工具后，追加 role=tool 消息（携带 tool_call_id）发起下一轮请求，模型会基于工具结果继续回答：

{
  "model": "deepseek-v4-flash",
  "messages": [
    { "role": "user", "content": "查询北京的天气" },
    {
      "role": "assistant",
      "content": null,
      "tool_calls": [
        { "id": "call_xxxx", "type": "function", "function": { "name": "get_weather", "arguments": "{\"city\": \"北京\"}" } }
      ]
    },
    { "role": "tool", "tool_call_id": "call_xxxx", "content": "北京今日晴，气温 20~28 度" }
  ]
}

多模态输入示例

以下为多模态（图文）对话请求格式，使用 OpenAI 兼容的 content 数组：

POST /v1/chat/completions
Authorization: Bearer sk-xxxxxxxx
Content-Type: application/json

{
  "model": "qwen3-vl-plus",
  "messages": [
    {
      "role": "user",
      "content": [
        { "type": "text", "text": "请描述这张图片的内容" },
        { "type": "image_url", "image_url": { "url": "https://example.com/pic.jpg" } }
      ]
    }
  ]
}

支持视频、音频输入的多模态模型（如 qwen3.5-omni-plus、qwen3.7-plus）可使用：

{
  "model": "qwen3.5-omni-plus",
  "messages": [
    {
      "role": "user",
      "content": [
        { "type": "text", "text": "帮我分析这段视频" },
        { "type": "video_url", "video_url": { "url": "https://example.com/video.mp4" } },
        { "type": "input_audio", "input_audio": { "data": "base64...", "format": "mp3" } }
      ]
    }
  ]
}

支持模型

以下为常用推荐模型，完整清单与单价以控制台实时报价为准：

计费说明

错误返回

接口统一返回 { code, msg } 结构（FastAdmin 规范），code 为 0 表示通用错误、非 0 表示对应业务错误：

{ "code": 0, "msg": "model is required" }
{ "code": 0, "msg": "messages is required" }
{ "code": 400, "msg": "messages 不能为空" }
{ "code": 402, "msg": "Insufficient balance" }
{ "code": 404, "msg": "Model not found" }

Token 用量与计费

最佳实践