Alibaba Cloud Model Studio：Anthropic API 相容

macOS

1. 在終端中執行以下命令，查看預設 Shell 類型。

   echo $SHELL

2. 根據 Shell 類型設定環境變數，命令如下：

   Zsh

   # 用百鍊 API KEY 替換 YOUR_DASHSCOPE_API_KEY
   echo 'export ANTHROPIC_BASE_URL="https://dashscope-intl.aliyuncs.com/apps/anthropic"' >> ~/.zshrc
   echo 'export ANTHROPIC_API_KEY="YOUR_DASHSCOPE_API_KEY"' >> ~/.zshrc

   Bash

   # 用百鍊 API Key 替換 YOUR_DASHSCOPE_API_KEY
   echo 'export ANTHROPIC_BASE_URL="https://dashscope-intl.aliyuncs.com/apps/anthropic"' >> ~/.bash_profile
   echo 'export ANTHROPIC_API_KEY="YOUR_DASHSCOPE_API_KEY"' >> ~/.bash_profile

3. 在終端中執行下列命令，使環境變數生效。

   Zsh

   source ~/.zshrc

   Bash

   source ~/.bash_profile

4. 開啟一個新的終端，執行下列命令，查看環境變數是否生效。

   echo $ANTHROPIC_BASE_URL
   echo $ANTHROPIC_API_KEY

Windows

1. 在 Windows 中，可以通過 CMD 或 PowerShell 將阿里雲百鍊提供的 Base URL 和API Key設定為環境變數。

   CMD

   1. 在 CMD 中運行以下命令，設定環境變數。

      # 用百鍊 API Key 替換 YOUR_DASHSCOPE_API_KEY
      setx ANTHROPIC_API_KEY "YOUR_DASHSCOPE_API_KEY"
      setx ANTHROPIC_BASE_URL "https://dashscope-intl.aliyuncs.com/apps/anthropic"

   2. 開啟一個新的 CMD 視窗，運行以下命令，檢查環境變數是否生效。

      echo %ANTHROPIC_API_KEY%
      echo %ANTHROPIC_BASE_URL%

   PowerShell

   1. 在 PowerShell 中運行以下命令，設定環境變數。

      # 用百鍊 API Key 替換 YOUR_DASHSCOPE_API_KEY
      [Environment]::SetEnvironmentVariable("ANTHROPIC_API_KEY", "YOUR_DASHSCOPE_API_KEY", [EnvironmentVariableTarget]::User)
      [Environment]::SetEnvironmentVariable("ANTHROPIC_BASE_URL", "https://dashscope-intl.aliyuncs.com/apps/anthropic", [EnvironmentVariableTarget]::User)

   2. 開啟一個新的 PowerShell 視窗，運行以下命令，檢查環境變數是否生效。

      echo $env:ANTHROPIC_API_KEY
      echo $env:ANTHROPIC_BASE_URL

cURL

curl -X POST "https://dashscope-intl.aliyuncs.com/apps/anthropic/v1/messages" \
  -H "Content-Type: application/json" \
  -H "x-api-key: ${ANTHROPIC_API_KEY}" \
  -d '{
    "model": "qwen-plus",
    "max_tokens": 1024,
    "stream": true,
    "thinking": {
      "type": "enabled",
      "budget_tokens": 1024
    },
    "system": "You are a helpful assistant",
    "messages": [
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": "你是誰？"
                }
            ]
        }
    ]
}'

Python

1. 安裝 Anthropic SDK

   pip install anthropic

2. 程式碼範例

   import anthropic
   import os
   
   client = anthropic.Anthropic(
       api_key=os.getenv("ANTHROPIC_API_KEY"),
       base_url=os.getenv("ANTHROPIC_BASE_URL"),
   )
   
   message = client.messages.create(
       model="qwen-plus",
       max_tokens=1024,
       stream=True,
       system="you are a helpful assistant",
       # 深度思考，僅支援部分模型，請參閱支援的模型列表
       thinking={
           "type": "enabled",
           "budget_tokens": 1024
       },
       messages=[
           {
               "role": "user",
               "content": [
                   {
                       "type": "text",
                       "text": "你是誰？"
                   }
               ]
           }
       ]
   )
   
   print("=== 思考過程 ===")
   first_text = True
   for chunk in message:
       if chunk.type == "content_block_delta":
           if hasattr(chunk.delta, 'thinking'):
               print(chunk.delta.thinking, end="", flush=True)
           elif hasattr(chunk.delta, 'text'):
               if first_text:
                   print("\n\n=== 回答 ===")
                   first_text = False
               print(chunk.delta.text, end="", flush=True)
   

TypeScript

1. 安裝 Anthropic TypeScript SDK

   npm install @anthropic-ai/sdk

2. 程式碼範例

   import Anthropic from "@anthropic-ai/sdk";
   
   async function main() {
     const anthropic = new Anthropic({
       apiKey: process.env.ANTHROPIC_API_KEY,
       baseURL: process.env.ANTHROPIC_BASE_URL,
     });
   
     const stream = await anthropic.messages.create({
       model: "qwen-plus",
       max_tokens: 1024,
       stream: true,
       // 深度思考，僅支援部分模型，請參閱支援的模型列表
       thinking: { type: "enabled", budget_tokens: 1024 },
       system: "You are a helpful assistant",
       messages: [{ 
         role: "user", 
         content: [
           {
             type: "text",
             text: "你是誰？"
           }
         ]
       }]
     });
   
     console.log("=== 思考過程 ===");
     let firstText = true;
   
     for await (const chunk of stream) {
       if (chunk.type === "content_block_delta") {
         if ('thinking' in chunk.delta) {
           process.stdout.write(chunk.delta.thinking);
         } else if ('text' in chunk.delta) {
           if (firstText) {
             console.log("\n\n=== 回答 ===");
             firstText = false;
           }
           process.stdout.write(chunk.delta.text);
         }
       }
     }
     console.log();
   }
   
   main().catch(console.error);