🚀 Light Browser Use Agent

2x Faster Browser Automation with 50% Less Token Consumption

A lightweight, optimized code agent that revolutionizes browser automation through simplified input/output design.

🔄 How It Works

The agent simplifies the traditional browser automation pipeline:

Core Innovation:

• Input: Highlighted screenshot only
• Output: Think + Python code for browser operations and data extraction

🚀 Quick Start

1. Setup Environment

# Copy and configure environment variables
cp .env.example .env
# Edit .env with your LLM API keys

2. Install Dependencies

# Install uv package manager
pip install uv

# Install project dependencies
uv sync

3. Run the Agent

from dotenv import load_dotenv
import asyncio
import os
load_dotenv()

from light_browser_use.agents.light_browser_use_agent import LightBrowserUseAgent
from light_browser_use.llm.openai_compatible import ChatOpenAICompatible

async def main():
    task = "search vibesurf, go into its github repo, summary the latest release for me"
    llm = ChatOpenAICompatible(model='qwen-vl-max',
                               base_url=os.getenv("DASHSCOPE_ENDPOINT"),
                               api_key=os.getenv("DASHSCOPE_API_KEY"))

    agent = LightBrowserUseAgent(
        task=task,
        max_steps=30,
        llm=llm
    )

    try:
        print('Running light browser agent...')
        session = await agent.run()
    finally:
        await agent.close()

asyncio.run(main())