browser-agent

What it does

Vision-based browser control from natural language: it sees the interface, plans actions, clicks and types precisely, extracts structured data and ships a test runner with visual assertions. Working from pixels rather than selectors is why it survives redesigns and odd custom widgets. Two constraints worth respecting — target sites' terms on automation, and a human confirmation step before anything irreversible. Screenshots of whatever is on screen reach your model provider. Apache-2.0.

Summary based on the project's own description. The Japanese page has a fuller beginner guide: setup difficulty, what to prepare, and a step-by-step start.

Screenshots & visuals

Before you start

• Setup difficultyFor advanced users
• Command lineRequired
• API keyRequired
• Sends data onlineYes

Judged automatically from the project's own metadata. Always confirm on the official page.

browser-agent FAQ

Is browser-agent free?

browser-agent itself is free, but it needs an API key for an external AI service. That service charges separately, though some have a free tier.

Do I need to install anything or write code to use browser-agent?

It needs the command line. If you are not comfortable typing commands, expect to get stuck during setup.

Does browser-agent work on a phone?

No. It needs a computer.

Can I use browser-agent commercially?

The licence is Apache-2.0, which permits commercial use. Conditions such as keeping the copyright notice still apply, so read the licence text before shipping it inside something.

How it compares

The conditions that decide whether you can actually use it, side by side.

"Sends data out" means your input is uploaded to someone's server. Pick "No" when the material cannot leave your organisation.

Other browser automation tools