Browser and Computer Use Agents

Open-source GUI automation framework for building desktop and browser agents with vision.

AI-powered web scraping and automation with a semantic query language for page elements.

Open-source desktop agent automating lead generation, knowledge library RAG, outreach, and scheduled workflows.

Drives a real Firefox from plain-English instructions, with sixteen MCP tools and a standalone web UI.

Enterprise-grade cloud browser infrastructure for AI agent automation at scale.

Open-source framework to let LLMs navigate and interact with any website programmatically.

Scalable headless browser infrastructure purpose-built for running AI agents in production.

AI-native browser from Atlassian and the Browser Company with built-in agent workflows.

Transparent agentic browser with visual workflow editing and persistent agentic memory.

On-device AI browser with 169+ local models that runs without internet connection.

Upcoming xAI desktop agent with mouse control and application automation capabilities.

Large Action Model framework to turn natural language instructions into browser automation.

Autonomous digital employee with Browser Operator extension acquired by Meta for web-based task execution.

Reliable web automation API with CAPTCHA handling for building production browser agents.

ChatGPT autonomous web agent with human checkpoints using Computer Use Agent technology.

Open-source Windows desktop agent with native computer-use over the Chrome DevTools Protocol, in-process cognition tools, and peer-to-peer coordination between instances.

Autonomous open-source computer use and browser automation agent framework.

MCP server wrapping Playwright for seamless browser automation integration with LLM agents.

Production browser automation with 21.5K GitHub stars, 10M+ executed workflows, vision-first automation, and sub-300ms response times.

AI web browsing framework built on Playwright with natural-language selectors and actions.

Task-list agent that runs on your own machine and completes tasks through your shell, files and logged-in browser, with any model.