Agent Browser
Overview
Open-source CLI that lets AI agents control Chrome with compact text output for browser automation tasks.
Agent Browser is Vercel’s Apache-2.0 browser-automation CLI. It provides navigation, compact page snapshots, clicks, form entry and screenshots for scripts or agents. A chat interface also lets a user give browsing instructions through a connected model.
Developers install the CLI and browser, provide authentication and decide which actions are permitted. Remote browser connections and experimental MCP support extend the setup. It fits testing and browser-based workflows, with human review for submissions, purchases or changes that matter.
Product overview
Capabilities, pricing and fit
Agent Browser provides browser commands for scripts and agents, plus a chat interface for model-driven browsing tasks.
Information as of August 8, 2026
What it can do
- Agent-oriented browser control
The CLI opens pages, returns compact page structure and stable element references, and supports interaction commands for clicking, typing, navigation, tabs, downloads, and screenshots.[1]
- Persistent and isolated sessions
Named sessions isolate browser instances, cookies, and storage, while state and profile options support restoring authenticated workflows.[1]
- Expanded automation surfaces
Recent releases add MCP tools, plugin support, hosted-sandbox helpers, network and HAR inspection, accessibility audits, and agent-readable page extraction.[2]
Where it fits best
- AI coding agents and automation scripts that need a command-line interface for inspecting and controlling browser sessions.[1]
- Local, sandboxed, or remote-browser workflows that need reusable sessions, compact accessibility references, downloads, screenshots, network capture, or an MCP tool surface.[1], [2]
Pricing and availability
Agent Browser is free and open source. Optional remote browser providers, hosted sandboxes, or external model services may have their own costs.[1]
Platforms: npm, Homebrew, Cargo, macOS, Linux, Windows[1], [2]
Access: Local CLI with a compatible browser, Remote browser connection, Model-driven chat with configured model access[1], [2]
Sources (2)
- vercel-labs/agent-browser | GitHubGitHub · checked 2026-10-12
- agent-browser CHANGELOG | GitHubChangelog · checked 2026-08-08
Autonomy level
45/100
Our subjective rating of how independently the product carries out its documented tasks, considering the human setup, input, approval and review it needs. This is not a measured success rate.
Comparisons
Custom Comparisons
Some of the use cases of Agent Browser:
- Adding browser automation capabilities to agent workflows
- Letting AI systems inspect and interact with web pages from a CLI
- Running repeatable browser tasks in development or automation environments
- Reducing browser state into compact text output for model context
- Building local browser-based automation workflows with Chrome
Loading Community Opinions...
Pricing model:
Code access:
Popularity level: 51%
Agent Browser Video:
Turn this profile into a trust signal
Show prospects that Agent Browser has a public place where they can check product details, pricing, ratings, and reviews.
Build confidence
Give buyers a third-party profile to explore.
Reduce hesitation
Put validation beside your strongest CTA.
Earn discovery
Every badge links prospects to your listing.
Choose your style
Preview it, then copy the complete embed code.
Shows buyers where to validate your product, pricing, and reputation.
Plain HTML. No signup, script, or maintenance required.
AI Agent Store · Specialist AI agentsFor browser automation developers
Your AI agent for Chrome automation delivery.
Exploring Agent Browser? For browser automation developers, your own AI agent can reproduce browser failures, build resilient web journeys, and verify completed runs. Start with a setup tailored to this work.
- ✓ Reproduce browser failures
- ✓ Build resilient web journeys
- ✓ Verify completed runs
Tailored starting setup. Chat on our site. See the workflow and monthly plan.