---
description: "与 OpenAI 完全兼容的 Chat Completions 文本对话接口，支持全部官方参数。"
---

# Chat Completions

> 端点: `POST https://oneapi.zaiwenai.com/v1/chat/completions`

与 OpenAI Chat Completions 接口完全兼容，支持所有 OpenAI API 参数。

## 请求参数

| 参数 | 类型 | 必填 | 说明 |
|------|------|------|------|
| `model` | string | 是 | 模型名称 |
| `messages` | array | 是 | 消息列表 |
| `stream` | boolean | 否 | 是否启用流式响应，默认 `false` |
| `temperature` | number | 否 | 采样温度，0-2 之间 |
| `max_tokens` | integer | 否 | 最大输出 Token 数 |
| `tools` | array | 否 | 工具定义列表，用于函数调用 |
| `tool_choice` | string \| object | 否 | 工具调用策略 |

其他参数请参考 [OpenAI Chat Completions 文档](https://platform.openai.com/docs/api-reference/chat)。

## Curl 示例

```bash
curl https://oneapi.zaiwenai.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $YOUR_API_KEY" \
  -d '{
    "model": "deepseek-v4-flash",
    "messages": [
      {
        "role": "system",
        "content": "You are a helpful assistant."
      },
      {
        "role": "user",
        "content": "Hello!"
      }
    ]
  }'
```

## OpenAI Python SDK 示例

```python
from openai import OpenAI

client = OpenAI(
    base_url="https://oneapi.zaiwenai.com/v1",
    api_key="$YOUR_API_KEY",
)

completion = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[
        {
            "role": "system",
            "content": "You are a helpful assistant.",
        },
        {
            "role": "user",
            "content": "hello!",
        },
    ],
)

print(completion.choices[0].message.content)
```

## 流式响应

```python
stream = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[
        {
            "role": "user",
            "content": "Tell me a short story",
        },
    ],
    stream=True,
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")
```

## 工具调用示例

```python
import json

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "获取指定城市的天气信息",
            "parameters": {
                "type": "object",
                "properties": {
                    "city": {
                        "type": "string",
                        "description": "城市名称",
                    },
                },
                "required": ["city"],
            },
        },
    },
]

completion = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[
        {
            "role": "user",
            "content": "北京今天天气怎么样？",
        },
    ],
    tools=tools,
    tool_choice="auto",
)

message = completion.choices[0].message
if message.tool_calls:
    for tool_call in message.tool_calls:
        print(f"调用工具: {tool_call.function.name}")
        print(f"参数: {tool_call.function.arguments}")
```

## LangChain 集成示例

```python
from pprint import pprint

from langchain.agents import create_agent
from langchain_openai import ChatOpenAI

model = ChatOpenAI(
    base_url="https://oneapi.zaiwenai.com/v1",
    api_key="$YOUR_API_KEY",
    model="deepseek-v4-flash",
)

_weather_db = {
    "Beijing": "Hazy, 28°C",
}


def get_user_city() -> str:
    """
    Asks the user for their current city via command-line input.
    """
    print("+++++ TOOL USER ++++: try to get user's city")
    return "Beijing"


def get_weather(city: str) -> str:
    """
    get weather for a given city from a mock database.
    """
    print(f"+++++ TOOL USER ++++: try to get weather for {city}")
    return _weather_db.get(
        city, f"Sorry, I don't have weather information for {city}."
    )


agent = create_agent(
    model=model,
    tools=[get_weather, get_user_city],
    system_prompt="You are a helpful assistant.",
)

r = agent.invoke(
    {"messages": [{"role": "user", "content": "What's the weather like?"}]}
)

pprint(r)
```

## 计费说明

按 Token 用量计费，支持缓存命中优惠。请求前会预估费用，余额不足时返回 402。