HomeCompaniesBrowser Use

Browser Use

Leading open-source web agent project with 50k stars in 3 months

We enable AI to control the browser. Browser-use is the biggest open-source web agent project. We got 40k stars in the last 3 months.
Browser Use
Founded:2024
Team Size:2
Status:
Active
Location:San Francisco
Group Partner:Jared Friedman
Active Founders

Magnus Müller, Founder

CEO of browser-use. Repeated founder. Scraping hobbyist + ML researcher. I have published for Cambridge CARES, studied Data Science at ETH Zurich, and worked in R&D for process automation companies. Since I learned to code, I have developed bots and web automations.
Magnus Müller
Magnus Müller
Browser Use

Gregor Zunic, Founder

I love solving hard problems. Data Science masters at ETH Zurich, bachelors in Physics.
Gregor Zunic
Gregor Zunic
Browser Use
Company Launches
Browser Use: Open Source alternative to OpenAI Operator
See original launch post ›

Hey everyone, we're Gregor and Magnus we are excited to launch Browser Use Cloud!

What is Browser Use (Cloud)?

We started this project 3 months ago because it simply wasn’t possible to use the Browser with any LLM. We thought “How hard could it be to build the interface between LLMs and the web?” Fast forwarding 3 months, we have combined 25k ⭐ on Github and our Open Source community is growing fast.

Four days ago, OpenAI launched Operator. Since then, our community has been asking non-stop for a hosted version of browser-use. So we shipped that. Now you can use Browser Use as a B2C chat tool at https://cloud.browser-use.com/. We are releasing the API in the next few days. If you need API access for web agents, call us, don’t wait for the Open AI Operator API to be released!

Existing problems

During many conversations with our discord community of 5000 people, we identified these core challenges in browser automation:

  • Websites change, and this breaks automation scripts. You usually only notice this, when it's too late.
  • Websites often catch bots. To avoid this, you need to use different IP addresses, solve captchas, and act more like a human.
  • Many people get frustrated with rate limits, parsing errors and API key management, when they do LLM-based scraping.
  • The most useful tasks often require you to log in. This can be a pain, especially if you have to type in your username and password every time.

Browser Use hosted version addresses these challenges by handling proxy rotation, maintaining persistent sessions - so you only need to login once, and allowing you to run as many instances in parallel as you need. We've priced it at $30/month, significantly lower than Operator.

On the open-source side, browser-use remains free and flexible. You can use any LLM you prefer - from Gemini to Sonnet, Qwen, and even DeepSeek-R1. It comes with an MIT license and full customisation options.

This week, we're working on follow-up questions, task reruns, voice mode, and scheduled agent starts. If you want to build automations directly into your products call us, we are happy to explore your use cases and help you set get set up with the API or even custom integration.

Our vision is simple: tell your computer what to do, and it gets it done.

We'd love to hear from you. What automation challenges are you facing? How do you think people will interact with computers in the future?