Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.browserbase.com/llms.txt

Use this file to discover all available pages before exploring further.

Agent Browser is a headless browser automation CLI built for AI agents. It provides a fast Rust-powered command-line interface with accessibility-first semantics, making it ideal for LLM-driven browser control. Agent Browser has built-in Browserbase support via its -p browserbase provider flag, giving your agents cloud browser sessions with no local browser required.

​
Key features

  • Accessibility tree snapshots with element refs (@e1, @e2) optimized for LLM consumption
  • Full browser control via simple CLI commands — click, type, fill, screenshot, navigate
  • Built-in Browserbase provider — connect to cloud browsers with a single flag
  • Rust-powered performance with sub-millisecond parsing overhead
  • Semantic locators — find elements by ARIA role, text, label, placeholder, or test ID

​
How it works

AI Agent → agent-browser CLI → Browserbase Cloud Browser → Target Website → Data Back
Your AI agent calls agent-browser commands to control a Browserbase cloud browser. The snapshot command returns an accessibility tree with element references that the agent can use for deterministic interactions.

​
When to use this

  • AI agent tooling — give LLMs a CLI interface to control browsers
  • Automated data extraction — extract data using accessibility tree snapshots
  • Form filling and testing — interact with web forms via simple commands
  • Screenshot and PDF generation — capture pages from cloud browsers

​
Get started

Quickstart guide

Set up Agent Browser with Browserbase and start automating in minutes.