Browser-Use API: Real-Time AI-Powered Browser Automation

A powerful and self-hosted REST API wrapper, aligning with the Browser-Use Cloud API v2 specification, designed to facilitate seamless browser automation with AI capabilities. This project offers a standardized REST interface empowering developers to integrate browser automation into their workflows efficiently.

Key Features

• Real-Time VNC Streaming: Monitor AI actions live via browser VNC streaming, enabling enhanced debugging and understanding of automation processes.
• Persistent Sessions: Manage browser profiles with saved login states to emulate human-like behavior across multiple tasks.
• Full Control of Automation Tasks: Start, stop, pause, and resume tasks at will, receiving instant updates as processes evolve.
• Diverse LLM Compatibility: Choose from over 15 language models, including GPT-4, Claude, and Gemini, optimizing performance based on project needs.
• Complete Data Control: Fully self-hosted to offer enhanced privacy and independence from cloud services, enabling offline operations.
• Production-Ready: Docker compatibility ensures smooth deployment, with multi-architecture support, health checks, and CI/CD readiness.

Usage Example

Begin utilizing the API with a simple command:

# Start the API service
docker run -d -p 8080:8080 reqeique/browser-use-api:latest

# Create a browser automation task
curl -X POST "http://localhost:8080/tasks" \
  -H "Content-Type: application/json" \
  -d '{"task": "Find the top 3 AI news on Hacker News", "llm": "gemini-flash-lite-latest"}'

# View the task in real-time

Advanced Use Cases

The API's flexibility supports various applications such as:

• Research Automation: Automate the search for up-to-date academic articles.
• E-commerce Monitoring: Track product prices across online platforms.
• Data Collection: Extract job listings or any other structured data from websites.
• Quality Assurance Testing: Execute end-to-end tests for web applications.

Configuration and Customization

Tasks can be customized using a comprehensive schema, allowing tailored requests for browser automation, including the ability to save session data and manage structured outputs.

Learning and Support

Extensive documentation provides guidance on setup, configuration, and troubleshooting. Engage with the community on Discord for support or to share ideas.