browser-use
🌐 Make websites accessible for AI agents. Automate tasks online with ease.
TL;DR · 30-second scan
browser-use (Python) — 🌐 Make websites accessible for AI agents. Automate tasks online with ease.
You need to automate a web flow that has no API and you can tolerate occasional failures and token cost.
AI & ML · Automation & Agents
Gives an LLM a browser and lets it click. You describe a task in English, it reads the page, decides on actions and drives Playwright underneath. As a way to automate flows that never had an API — internal dashboards, vendor portals, government sites — it is the most practical version of this idea so far. It went from nothing to six figures of stars faster than almost anything in the AI tooling space. The tradeoff is cost and reliability: every step is a model call, so a long task gets expensive, and pages that change layout break runs in ways a scripted scraper would not. Good for exploration, risky as unattended production infrastructure.
You need to automate a web flow that has no API and you can tolerate occasional failures and token cost.
The site has an API, or the task must run unattended and identically every time — write a normal Playwright script.
Add this badge to your README to show your project is curated on StackPicks. Free, lightweight (180×28 SVG), and gives your visitors a one-click way to see honest take + alternatives.
[](https://stackpicks.dev/repo/browser-use-browser-use)
<a href="https://stackpicks.dev/repo/browser-use-browser-use"><img src="https://stackpicks.dev/api/badge/browser-use-browser-use" alt="Featured on StackPicks" width="180" height="28" /></a>
Are you the maintainer of browser-use/browser-use? Add the badge and we'll feature your project in the weekly curator newsletter.