Chat Completions

用于多轮回话、文本生成、摘要、信息抽取和结构化助手场景。

Endpoint

POST https://pinducloud.cc/v1/chat/completions

请求示例

curl https://pinducloud.cc/v1/chat/completions \
  -H "Authorization: Bearer $PINDUCLOUD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5",
    "messages": [
      {
        "role": "system",
        "content": "You are a precise API documentation assistant."
      },
      {
        "role": "user",
        "content": "List three production API security practices."
      }
    ],
    "temperature": 0.3,
    "max_tokens": 500
  }'

响应示例

{
  "id": "chatcmpl_abc123",
  "object": "chat.completion",
  "created": 1710000000,
  "model": "gpt-5.5",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "1. Keep API keys on trusted servers.\n2. Apply request timeouts and retry limits.\n3. Log errors with sensitive data removed."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 38,
    "completion_tokens": 42,
    "total_tokens": 80
  }
}

流式请求

curl https://pinducloud.cc/v1/chat/completions \
  -H "Authorization: Bearer $PINDUCLOUD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5",
    "messages": [
      {
        "role": "user",
        "content": "Create a short launch checklist."
      }
    ],
    "stream": true
  }'

参数

参数	类型	必填	说明
model	string	是	模型 ID，例如 gpt-5.5
messages	array	是	OpenAI 格式的回话消息
temperature	number	否	采样温度
max_tokens	number	否	最大输出 token 数
stream	boolean	否	是否返回流式响应

接入建议

• 默认低延迟场景可使用 gpt-5.5。
• Claude 系列模型请改用 /v1/messages，不要通过本接口调用。
• 服务端应设置请求超时和最大重试次数。