Chat Completions API

POST /v1/chat/completions 是最常用的端点，用于与语言模型进行多轮对话。

请求

POST https://real200.com/v1/chat/completions

Headers

Header	类型	必填	说明
Authorization	string	是	Bearer sk-real200-xxx
Content-Type	string	是	application/json

Body 参数

参数	类型	必填	默认值	说明
model	string	是	-	模型名称，如 gpt-4o、claude-sonnet-4
messages	array	是	-	对话消息列表
temperature	number	否	1.0	随机性，范围 0–2
top_p	number	否	1.0	核采样概率阈值
max_tokens	integer	否	模型默认	最大生成 Token 数
stream	boolean	否	false	是否启用流式输出
stop	string/array	否	null	停止序列
presence_penalty	number	否	0	存在惩罚，范围 -2–2
frequency_penalty	number	否	0	频率惩罚，范围 -2–2
response_format	object	否	-	指定 JSON 响应格式
tools	array	否	-	工具（函数调用）定义
tool_choice	string/object	否	"auto"	工具选择策略
seed	integer	否	-	随机种子（可复现）

messages 格式

{
  "messages": [
    {"role": "system", "content": "你是一个有用的助手。"},
    {"role": "user", "content": "今天天气怎么样？"},
    {"role": "assistant", "content": "我无法获取实时天气信息..."},
    {"role": "user", "content": "那北京的呢？"}
  ]
}

示例请求

curl https://real200.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $REAL200_API_KEY" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {"role": "user", "content": "请用三句话介绍你自己。"}
    ],
    "temperature": 0.7,
    "max_tokens": 200
  }'

响应

非流式响应

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1716300000,
  "model": "gpt-4o",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "我是一个 AI 助手..."
      },
      "finish_reason": "stop",
      "logprobs": null
    }
  ],
  "usage": {
    "prompt_tokens": 12,
    "completion_tokens": 45,
    "total_tokens": 57
  },
  "system_fingerprint": "fp_abc123"
}

流式响应

启用 stream: true 后，响应以 text/event-stream 格式返回：

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"我是"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"一个"},"finish_reason":null}]}

data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

finish_reason 说明

值	说明
stop	模型自然结束输出
length	达到 max_tokens 限制
content_filter	内容被安全过滤
tool_calls	模型调用了工具

错误码

HTTP 状态码	错误类型	说明
400	invalid_request_error	请求参数错误
401	invalid_api_key	API Key 无效
402	insufficient_quota	余额或额度不足
429	rate_limit_exceeded	请求频率超限
500	api_error	服务端错误

详见错误码字典。