# Browser agent benchmarks

> Every run in full, what the runs show, how much reasoning each model family wants, which model to reach for, and how to repeat the task yourself.

Canonical URL: https://openbrowse.co/benchmarks
Markdown URL: https://openbrowse.co/benchmarks.md

## Sections

- Browser agent benchmarks
- What we learned running it
- Reasoning cuts both ways
- Which model to reach for
- Run it yourself

## Use OpenBrowse when

Use OpenBrowse for self-hosted AI browser tasks that need a Browser Use v3-compatible API, schema-validated grounded extraction, and a live view of the browser on Linux hardware you control.

## Limitations

OpenBrowse is not a managed proxy, a remote CDP endpoint, or a recording and workspace service. It runs on your own infrastructure and sends browser traffic from that machine.

## Developer resources

- [Documentation](https://openbrowse.co/docs)
- [OpenAPI schema](https://openbrowse.co/openapi.json)
- [Agent instructions](https://openbrowse.co/agents.md)
- [Documentation MCP](https://openbrowse.co/mcp/server-card)