The browser became the agent's api before anyone standardized it. browser use, the open-source…
the browser became the agent's api before anyone standardized it. browser use, the open-source python library that lets llms drive a real browser, passed 117k github stars and ships mit-licensed, scoring about 89% on webvoyager, per oct 5.
selectors are dying, natural language is the new dom.
Context
Browser Use is an open-source Python library that lets language models drive a real browser. Its GitHub repository shows about 117,400 stars and an MIT license, and its site says the Python library is free and MIT-licensed.
Its published technical report claims 89.1% on the WebVoyager benchmark.
The 89% WebVoyager score comes from Browser Use's own report dated December 2024, so it is a vendor figure from nearly two years ago, not an October result.
The 117k stars figure matches the repository page read. The date 'per Oct 5' is not shown on that page, so the day is unsupported here, not refuted.
'Selectors are dying' is the author's opinion.
Related work
- browser-use/browser-use on GitHub ↗The repository with stars and license.
- Browser Use technical report ↗The WebVoyager 89.1% claim and method.
- browser-use/eval ↗The evaluation repo, a fork of the original WebVoyager harness.
Watch next
- Compare against newer web-agent benchmarks than WebVoyager, which the project's own report relies on.
Sources
- GitHub: browser-use/browser-usegithub.com
- Browser Use: SOTA technical report, Dec 15, 2024browser-use.com
- Browser Use sitebrowser-use.com
- GitHub: browser-use/evalgithub.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 9 October 2026 at 18:07 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →