聊天接口 (Chat)

约 962 字大约 3 分钟

2025-02-10

聊天接口是最常用的 AI 接口，支持多轮对话、流式输出等功能。

接口地址

POST https://api.ai.webweb.cn/v1/chat/completions

请求参数

必填参数

参数	类型	说明
model	string	模型名称，如 `gpt-3.5-turbo`、`gpt-4`
messages	array	对话消息数组

可选参数

参数	类型	默认值	说明
temperature	number	1	温度参数，控制随机性，范围 0-2
top_p	number	1	核采样参数
n	integer	1	生成的回复数量
stream	boolean	false	是否使用流式输出
max_tokens	integer	-	最大生成 Token 数
presence_penalty	number	0	存在惩罚，范围 -2.0 到 2.0
frequency_penalty	number	0	频率惩罚，范围 -2.0 到 2.0
stop	string/array	null	停止词
user	string	-	用户标识符

消息格式

每条消息包含以下字段：

字段	类型	说明
role	string	角色：`system`、`user`、`assistant`
content	string	消息内容

请求示例

基础请求

cURL

curl https://api.ai.webweb.cn/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -d '{
    "model": "gpt-3.5-turbo",
    "messages": [
      {"role": "system", "content": "你是一个有帮助的助手。"},
      {"role": "user", "content": "你好，请介绍一下你自己。"}
    ]
  }'

Python

from openai import OpenAI

client = OpenAI(
    api_key="sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
    base_url="https://api.ai.webweb.cn/v1"
)

response = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "system", "content": "你是一个有帮助的助手。"},
        {"role": "user", "content": "你好，请介绍一下你自己。"}
    ]
)

print(response.choices[0].message.content)

JavaScript

import OpenAI from 'openai';

const openai = new OpenAI({
    apiKey: 'sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx',
    baseURL: 'https://api.ai.webweb.cn/v1'
});

const response = await openai.chat.completions.create({
    model: 'gpt-3.5-turbo',
    messages: [
        { role: 'system', content: '你是一个有帮助的助手。' },
        { role: 'user', content: '你好，请介绍一下你自己。' }
    ]
});

console.log(response.choices[0].message.content);

package main

import (
    "context"
    "fmt"
    openai "github.com/sashabaranov/go-openai"
)

func main() {
    config := openai.DefaultConfig("sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx")
    config.BaseURL = "https://api.ai.webweb.cn/v1"
    client := openai.NewClientWithConfig(config)

    resp, err := client.CreateChatCompletion(
        context.Background(),
        openai.ChatCompletionRequest{
            Model: openai.GPT3Dot5Turbo,
            Messages: []openai.ChatCompletionMessage{
                {Role: openai.ChatMessageRoleSystem, Content: "你是一个有帮助的助手。"},
                {Role: openai.ChatMessageRoleUser, Content: "你好，请介绍一下你自己。"},
            },
        },
    )

    if err != nil {
        fmt.Printf("Error: %v\n", err)
        return
    }

    fmt.Println(resp.Choices[0].Message.Content)
}

流式输出

Python

from openai import OpenAI

client = OpenAI(
    api_key="sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
    base_url="https://api.ai.webweb.cn/v1"
)

stream = client.chat.completions.create(
    model="gpt-3.5-turbo",
    messages=[
        {"role": "user", "content": "写一首关于春天的诗"}
    ],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content is not None:
        print(chunk.choices[0].delta.content, end="")

JavaScript

import OpenAI from 'openai';

const openai = new OpenAI({
    apiKey: 'sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx',
    baseURL: 'https://api.ai.webweb.cn/v1'
});

const stream = await openai.chat.completions.create({
    model: 'gpt-3.5-turbo',
    messages: [{ role: 'user', content: '写一首关于春天的诗' }],
    stream: true
});

for await (const chunk of stream) {
    process.stdout.write(chunk.choices[0]?.delta?.content || '');
}

响应示例

非流式响应

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1677858242,
  "model": "gpt-3.5-turbo-0613",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "你好！我是一个 AI 助手，很高兴为您服务。"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 25,
    "completion_tokens": 20,
    "total_tokens": 45
  }
}

流式响应

{"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}
{"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{"content":"你"},"finish_reason":null}]}
{"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{"content":"好"},"finish_reason":null}]}
{"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-3.5-turbo-0613","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

支持的模型

模型	上下文长度	说明
gpt-4-turbo	128K	GPT-4 Turbo，最新最强
gpt-4	8K	GPT-4 标准版
gpt-4-32k	32K	GPT-4 长上下文版
gpt-3.5-turbo	16K	GPT-3.5 Turbo
claude-3-opus	200K	Claude 3 最强版
claude-3-sonnet	200K	Claude 3 平衡版
claude-3-haiku	200K	Claude 3 快速版
gemini-pro	32K	Google Gemini Pro
deepseek-chat	32K	Deepseek 聊天模型

最佳实践

1. 使用 System 消息

合理使用 system 消息可以更好地控制 AI 的行为：

messages = [
    {"role": "system", "content": "你是一个专业的Python程序员，只用中文回答。"},
    {"role": "user", "content": "如何读取一个JSON文件？"}
]

2. 管理对话历史

保持合理的对话历史长度，避免超出 Token 限制：

def trim_messages(messages, max_tokens=4000):
    """保留最近的消息，确保不超过 Token 限制"""
    # 实现截断逻辑
    pass

3. 错误处理

from openai import OpenAI, APIError, RateLimitError

client = OpenAI(base_url="https://api.ai.webweb.cn/v1")

try:
    response = client.chat.completions.create(
        model="gpt-3.5-turbo",
        messages=[{"role": "user", "content": "Hello!"}]
    )
except RateLimitError:
    print("请求过于频繁，请稍后重试")
except APIError as e:
    print(f"API 错误: {e}")

版权所有

许可证：署名 4.0 国际 (CC-BY-4.0)