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

# 快速开始

> 2 分钟内开始使用 ARouter。无需修改代码——只需替换 base_url 和 api_key。

ARouter 为主流 LLM 提供商提供统一的 API key 和单一端点。如果你已经在使用 OpenAI 兼容的客户端，迁移通常只需更改 `base_url`、`api_key`，以及可选的应用归因请求头。

## 1. 获取 API Key

在 [ARouter 控制台](https://api.arouter.ai) 注册并创建 API key。
你的 key 格式类似 `lr_live_xxxxxxxxxxxx`。

## 2. 安装 ARouter SDK

官方 `@arouter/sdk` 适用于任何 Node.js 或 TypeScript 项目，支持 npm、yarn 和 pnpm。

<Tabs>
  <Tab title="npm">
    ```bash theme={null}
    npm install @arouter/sdk
    ```
  </Tab>

  <Tab title="yarn">
    ```bash theme={null}
    yarn add @arouter/sdk
    ```
  </Tab>

  <Tab title="pnpm">
    ```bash theme={null}
    pnpm add @arouter/sdk
    ```
  </Tab>
</Tabs>

```typescript theme={null}
import { ARouter } from "@arouter/sdk";

const client = new ARouter({
  apiKey: "lr_live_xxxx",
  baseURL: "https://api.arouter.ai",
});

const response = await client.chatCompletion({
  model: "openai/gpt-5.4",
  messages: [{ role: "user", content: "你好！" }],
});
console.log(response.choices[0].message.content);
```

查看 [Node.js / TypeScript SDK 指南](/zh/sdks/node)，了解流式输出、key 管理和 x402 支付示例。

## 3. 使用现有 SDK

已在使用 OpenAI、Anthropic 或 Go？只需更改 `base_url` 和 `api_key` 即可。

<Tabs>
  <Tab title="Python (OpenAI)">
    ```python theme={null}
    from openai import OpenAI

    client = OpenAI(
        base_url="https://api.arouter.ai/v1",
        api_key="lr_live_xxxx",
        default_headers={
            "HTTP-Referer": "https://myapp.com",  # 可选
            "X-Title": "My AI App",               # 可选
        },
    )

    response = client.chat.completions.create(
        model="openai/gpt-5.4",
        messages=[{"role": "user", "content": "你好！"}],
    )
    print(response.choices[0].message.content)
    ```
  </Tab>

  <Tab title="Node.js (OpenAI)">
    ```typescript theme={null}
    import OpenAI from "openai";

    const client = new OpenAI({
      baseURL: "https://api.arouter.ai/v1",
      apiKey: "lr_live_xxxx",
      defaultHeaders: {
        "HTTP-Referer": "https://myapp.com", // 可选
        "X-Title": "My AI App",              // 可选
      },
    });

    const response = await client.chat.completions.create({
      model: "openai/gpt-5.4",
      messages: [{ role: "user", content: "你好！" }],
    });
    console.log(response.choices[0].message.content);
    ```
  </Tab>

  <Tab title="Python (Anthropic)">
    ```python theme={null}
    import anthropic

    client = anthropic.Anthropic(
        base_url="https://api.arouter.ai",
        api_key="lr_live_xxxx",
    )

    message = client.messages.create(
        model="claude-sonnet-4.6",
        max_tokens=1024,
        messages=[{"role": "user", "content": "你好！"}],
    )
    print(message.content[0].text)
    ```
  </Tab>

  <Tab title="Go">
    ```go theme={null}
    package main

    import (
        "context"
        "fmt"
        "log"

        "github.com/arouter-ai/arouter-go"
    )

    func main() {
        client := arouter.NewClient("lr_live_xxxx",
            arouter.WithBaseURL("https://api.arouter.ai/v1"),
            arouter.WithHeader("HTTP-Referer", "https://myapp.com"),
            arouter.WithHeader("X-Title", "My AI App"),
        )

        resp, err := client.CreateChatCompletion(context.Background(), arouter.ChatCompletionRequest{
            Model: "openai/gpt-5.4",
            Messages: []arouter.Message{
                {Role: "user", Content: "你好！"},
            },
        })
        if err != nil {
            log.Fatal(err)
        }

        fmt.Println(resp.Choices[0].Message.Content)
    }
    ```
  </Tab>

  <Tab title="cURL">
    ```bash theme={null}
    curl https://api.arouter.ai/v1/chat/completions \
      -H "Authorization: Bearer lr_live_xxxx" \
      -H "HTTP-Referer: https://myapp.com" \
      -H "X-Title: My AI App" \
      -H "Content-Type: application/json" \
      -d '{
        "model": "openai/gpt-5.4",
        "messages": [{"role": "user", "content": "你好！"}]
      }'
    ```
  </Tab>

  <Tab title="fetch">
    ```typescript theme={null}
    const response = await fetch('https://api.arouter.ai/v1/chat/completions', {
      method: 'POST',
      headers: {
        'Authorization': 'Bearer lr_live_xxxx',
        'Content-Type': 'application/json',
        'HTTP-Referer': 'https://myapp.com', // 可选：来源追踪
        'X-Title': 'My AI App',              // 可选：显示名称
      },
      body: JSON.stringify({
        model: 'openai/gpt-5.4',
        messages: [{ role: 'user', content: '你好！' }],
      }),
    });

    const data = await response.json();
    console.log(data.choices[0].message.content);
    ```
  </Tab>
</Tabs>

<Note>
  `HTTP-Referer` 和 `X-Title` 为可选项。如果你希望 ARouter 控制台分析将请求归因到特定应用或工作流，请包含这两个请求头。
</Note>

## 4. 直接调用 API

如果不想安装 SDK，可以使用任意 HTTP 客户端调用 ARouter：

```python theme={null}
import json
import requests

response = requests.post(
    "https://api.arouter.ai/v1/chat/completions",
    headers={
        "Authorization": "Bearer lr_live_xxxx",
        "HTTP-Referer": "https://myapp.com",  # 可选
        "X-Title": "My AI App",               # 可选
        "Content-Type": "application/json",
    },
    data=json.dumps({
        "model": "openai/gpt-5.4",
        "messages": [{"role": "user", "content": "你好！"}],
    }),
)

print(response.json()["choices"][0]["message"]["content"])
```

## 5. 切换不同提供商

使用 ARouter，切换提供商只需更改 `model` 字符串：

```python theme={null}
# OpenAI
response = client.chat.completions.create(model="openai/gpt-5.4", ...)

# Anthropic（通过 OpenAI SDK！）
response = client.chat.completions.create(model="anthropic/claude-sonnet-4.6", ...)

# Google Gemini
response = client.chat.completions.create(model="google/gemini-2.5-flash", ...)

# DeepSeek
response = client.chat.completions.create(model="deepseek/deepseek-v3.2", ...)
```

<Tip>
  如果省略提供商前缀（例如只写 `"gpt-5.4"`），ARouter 默认使用 OpenAI。
</Tip>

## 下一步

<CardGroup cols={2}>
  <Card title="身份验证" icon="lock" href="/zh/authentication">
    了解三种验证方式
  </Card>

  <Card title="模型路由" icon="shuffle" href="/zh/model-routing">
    理解 provider/model 格式与多模型路由
  </Card>

  <Card title="账单与积分" icon="credit-card" href="/zh/guides/billing-and-credits">
    查看定价、余额和积分规则
  </Card>

  <Card title="请求归因" icon="globe" href="/zh/guides/request-attribution">
    在控制台分析中将流量归因到你的应用
  </Card>

  <Card title="流式输出" icon="wave-sine" href="/zh/guides/streaming">
    启用实时流式响应
  </Card>

  <Card title="工具调用" icon="wrench" href="/zh/guides/features/tool-calling">
    让模型访问你的函数
  </Card>

  <Card title="结构化输出" icon="brackets-curly" href="/zh/guides/features/structured-outputs">
    强制模型返回有效的 JSON schema
  </Card>

  <Card title="提示词缓存" icon="bolt" href="/zh/guides/features/prompt-caching">
    降低重复提示的成本和延迟
  </Card>

  <Card title="Key 管理" icon="key" href="/zh/guides/key-management">
    为团队创建受限 key
  </Card>

  <Card title="常见问题" icon="circle-question" href="/zh/faq">
    常见问题解答
  </Card>
</CardGroup>