基于大语言模型的浏览器自动化：playwright+browser-use

 

 

1、安装nodejs

https://nodejs.org/zh-cn/download 下载 node-v24.11.0-x64.msi

2、安装 playwright

npx playwright install

 

 

3、安装 browser-use 包

首先需要安装了python环境，然后创建 python 虚拟环境

 创建 python 虚拟环境，也可以使用

conda create --name browser_use python=3.11
conda activate browser_use

安装 browser-use 包

pip install browser-use -i https://pypi.tuna.tsinghua.edu.cn/simple

 

4、根据使用的大模型服务，在环境变量设置里配置相应的 API key

OPENAI_API_KEY（gpt-4o 模型）
ANTHROPIC_API_KEY（claude 模型）
GEMINI_API_KEY（gemini 模型）
DEEPSEEK_API_KEY（deepseek 的 r1、v3 模型）

 

5、示例1：浏览器搜索，回答用户问题

需要用到的依赖模块

langchain-openai：提供ChatOpenAI类，用于与OpenAI的聊天模型进行交互
browser-use：包含Agent类，实现智能浏览器自动化功能
asyncio：Python标准库，用于异步编程，无需额外安装
os：Python标准库，提供操作系统接口，无需额外安装
pydantic：提供SecretStr等数据验证功能，用于安全处理敏感信息

安装依赖

pip install langchain-openai pydantic -i https://pypi.tuna.tsinghua.edu.cn/simple

 

 

安装browser-use环境

参考：https://github.com/browser-use/browser-use?tab=readme-ov-file

1. Create environment with uv (Python>=3.11):

uv init

2. Install Browser-Use package:

#  We ship every day - use the latest version!
uv add browser-use
uv sync

执行结果：

PS C:\Users\EDY> uv init
Initialized project `edy`
PS C:\Users\EDY> uv sync
>> uv add browser-use
>> #  We ship every day - use the latest version!
Using CPython 3.13.2 interpreter at: Desktop\work\installtools\python-3.13.2-amd64\python.exe
Creating virtual environment at: .venv
Resolved 1 package in 3ms
Audited in 0.02ms
Resolved 261 packages in 3.31s
Prepared 97 packages in 10.93s
Installed 97 packages in 433ms
 + aiofiles==25.1.0
 + aiohappyeyeballs==2.6.1
 + aiohttp==3.12.15
 + aiosignal==1.4.0
 + annotated-types==0.7.0
 + anthropic==0.72.0
 + anyio==4.11.0
 + attrs==25.4.0
 + authlib==1.6.5
 + backoff==2.2.1
 + beautifulsoup4==4.14.2
 + browser-use==0.9.4
 + bubus==1.5.6
 + cachetools==6.2.1
 + cdp-use==1.4.3
 + certifi==2025.10.5
 + cffi==2.0.0
 + charset-normalizer==3.4.4
 + click==8.3.0
 + colorama==0.4.6
 + cryptography==46.0.3
 + distro==1.9.0
 + docstring-parser==0.17.0
 + frozenlist==1.8.0
 + google-api-core==2.28.1
 + google-api-python-client==2.185.0
 + google-auth==2.42.0
 + google-auth-httplib2==0.2.0
 + google-auth-oauthlib==1.2.2
 + google-genai==1.46.0
 + googleapis-common-protos==1.71.0
 + groq==0.33.0
 + h11==0.16.0
 + httpcore==1.0.9
 + httplib2==0.31.0
 + httpx==0.28.1
 + httpx-sse==0.4.3
 + idna==3.11
 + inquirerpy==0.3.4
 + jiter==0.11.1
 + jsonschema==4.25.1
 + jsonschema-specifications==2025.9.1
 + markdown-it-py==4.0.0
 + markdownify==1.2.0
 + mcp==1.19.0
 + mdurl==0.1.2
 + multidict==6.7.0
 + oauthlib==3.3.1
 + ollama==0.6.0
 + openai==1.109.1
 + pfzy==0.3.4
 + pillow==12.0.0
 + portalocker==2.10.1
 + posthog==6.7.11
 + prompt-toolkit==3.0.52
 + propcache==0.4.1
 + proto-plus==1.26.1
 + protobuf==6.33.0
 + psutil==7.1.2
 + pyasn1==0.6.1
 + pyasn1-modules==0.4.2
 + pycparser==2.23
 + pydantic==2.12.3
 + pydantic-core==2.41.4
 + pydantic-settings==2.11.0
 + pygments==2.19.2
 + pyotp==2.9.0
 + pyparsing==3.2.5
 + pypdf==6.1.3
 + python-dateutil==2.9.0.post0
 + python-dotenv==1.2.1
 + python-multipart==0.0.20
 + pywin32==311
 + referencing==0.37.0
 + reportlab==4.4.4
 + requests==2.32.5
 + requests-oauthlib==2.0.0
 + rich==14.2.0
 + rpds-py==0.28.0
 + rsa==4.9.1
 + screeninfo==0.8.1
 + six==1.17.0
 + sniffio==1.3.1
 + soupsieve==2.8
 + sse-starlette==3.0.2
 + starlette==0.49.1
 + tenacity==9.1.2
 + tqdm==4.67.1
 + typing-extensions==4.15.0
 + typing-inspection==0.4.2
 + uritemplate==4.2.0
 + urllib3==2.5.0
 + uuid7==0.1.0
 + uvicorn==0.38.0
 + wcwidth==0.2.14
 + websockets==15.0.1
 + yarl==1.22.0
PS C:\Users\EDY>

3. Get your API key from Browser Use Cloud and add it to your .env file (new signups get $10 free credits):

# .env
BROWSER_USE_API_KEY=your-key

4. Install Chromium browser:

uvx browser-use install

执行结果：

 5. Run your first agent:

from browser_use import Agent, Browser, ChatBrowserUse
import asyncio

async def example():
    browser = Browser(
        # use_cloud=True,  # Uncomment to use a stealth browser on Browser Use Cloud
    )

    llm = ChatBrowserUse()

    agent = Agent(
        task="Find the number of stars of the browser-use repo",
        llm=llm,
        browser=browser,
    )

    history = await agent.run()
    return history

if __name__ == "__main__":
    history = asyncio.run(example())