> ## Documentation Index
> Fetch the complete documentation index at: https://docs.codeflare.cc/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat Completions

> 使用 messages 与 OpenAI 兼容聊天入口。

## 基础调用

入口为 `POST /v1/chat/completions`。模型需要支持 `chat-completions`；使用 Bearer 鉴权、`messages` 数组与合理的 `max_completion_tokens` 上限。

<CodeGroup>
  ```bash cURL theme={null}
  curl "https://YOUR-REGION.example/v1/chat/completions" \
    -H "Authorization: Bearer $CODEFLARE_API_KEY" \
    -H 'Content-Type: application/json' \
    -d '{"model":"MODEL_ID","messages":[{"role":"user","content":"Hello"}],"max_completion_tokens":256}'
  ```

  ```python Python theme={null}
  from openai import OpenAI
  import os

  client = OpenAI(
      api_key=os.environ["CODEFLARE_API_KEY"],
      base_url="https://YOUR-REGION.example/v1", max_retries=0,
  )
  response = client.chat.completions.create(
      model="MODEL_ID",
      messages=[{"role": "user", "content": "Hello"}],
      max_completion_tokens=256,
  )
  print(response.choices[0].message.content)
  ```

  ```javascript JavaScript theme={null}
  import OpenAI from "openai";
  const client = new OpenAI({
    apiKey: process.env.CODEFLARE_API_KEY,
    baseURL: "https://YOUR-REGION.example/v1", maxRetries: 0,
  });
  const response = await client.chat.completions.create({
    model: "MODEL_ID", messages: [{ role: "user", content: "Hello" }],
    max_completion_tokens: 256,
  });
  console.log(response.choices[0].message.content);
  ```
</CodeGroup>

## 流式输出

设置 `stream: true`。Python 使用 `for chunk in stream`，读取非空 `chunk.choices[0].delta.content`；JavaScript 使用 `for await` 读取 `choices[0]?.delta?.content`。HTTP 原生 SSE 在结束时包含 `[DONE]`，不要使用 Responses 的 `response.output_text.delta` 解析器。

## 选择协议

此入口与 Responses 使用不同的请求和输出格式。Codex CLI 使用 Responses；Claude Code 使用 Messages。若要保存、查询和续接平台响应，请查看 [Responses](/zh/api/responses)。

工具调用、结构化输出和多模态受模型及上游能力约束。自动重试应关闭；断连后先检查日志再发起新请求。


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.