Browser Automation

Puppeteer

A JavaScript library that controls Chrome and Firefox through a high-level API for automation, scraping, and testing.

Quick decision

Best for: JavaScript teams that want direct Chrome control, screenshots, PDFs, and page automation without a test-runner layer.
Not ideal for: Cross-engine work that needs WebKit, or teams that want a built-in test runner with assertions and tracing. That is Playwright territory.
Pricing model: Open source under Apache-2.0. The library is free to use. You provide the machine, the browser, and any infrastructure around it.
Deployment: Self-hosted library. Local browser.
Authentication: No account needed for the library. Sites you automate handle their own sign-in.
Review state: Source verified. Facts checked 2026-10-03. Not locally tested by AgentsUse.

What it does

Puppeteer is a JavaScript library maintained by the Chrome team at Google. It provides a high-level API to control Chrome or Firefox over the DevTools Protocol or WebDriver BiDi, and it runs headless by default.

Installing the main package downloads a compatible Chrome build. The puppeteer-core package installs only the library, for cases where you point Puppeteer at a browser you manage yourself.

For agent workflows, the project points readers to chrome-devtools-mcp, a Puppeteer-based MCP server for browser automation and debugging. Puppeteer also documents experimental WebMCP support.

Verified capabilities

Browser control

Chrome and Firefox control Source verified
Control Chrome or Firefox over the DevTools Protocol or WebDriver BiDi.
Page navigation and input Source verified
Open pages, set viewports, type, and click through locators.

Data access

Screenshots and page capture Source verified
Capture pages as part of scraping and QA workflows.

Setup

Bundled browser download Source verified
The main package downloads a compatible Chrome during installation, with a browsers install command when the script is blocked.
Library-only install Source verified
puppeteer-core installs without downloading Chrome, for bring-your-own-browser setups.

Agent integration

MCP path Source verified
The README points to chrome-devtools-mcp, a Puppeteer-based MCP server for automation and debugging.

Quick start

Install Puppeteer

From the official README. Modern package managers may block the browser download script, and the README documents the follow-up command.

npm i puppeteer
# if the browser download was blocked:
# npx puppeteer browsers install

Source: https://github.com/puppeteer/puppeteer. Examples use placeholders only. Never paste a real key into a profile, config file you share, or a ticket.

MCP support: Via a separate project. The Puppeteer README directs readers to chrome-devtools-mcp, a Puppeteer-based MCP server. AgentsUse has not published a separate MCP page for it yet.

Works with

MCP

Puppeteer-based MCP server (chrome-devtools-mcp) linked from the README.

Packages

npm: puppeteer and puppeteer-core.

Browsers

Chrome and Firefox through DevTools Protocol or WebDriver BiDi.

Deployment

Self-hosted library. You run the browsers.

Only sourced support is listed. A missing framework means AgentsUse has not verified it yet, not that it cannot work.

Health and maintenance

GitHub stars: 95,649 (checked 2026-10-03)
GitHub forks: 9,588 (checked 2026-10-03)
Package: npm: puppeteer
License: Open source, Apache-2.0
Maintainer: Google

Stars and forks from the GitHub repository page. No package version is shown because AgentsUse did not re-check the npm version field in this pass. Maintenance signals only, not a quality rating.

Pricing and license

Open source under Apache-2.0. The library is free to use. You provide the machine, the browser, and any infrastructure around it.

Limitations and safety

No WebKit engine. If you need Safari-engine coverage, look at Playwright instead.
Package managers that block install scripts will skip the Chrome download until you run the browsers install command.
No independently checked benchmark is published for this tool yet. The maintenance and setup facts above are source checked, but they do not measure task quality or speed.

Browser and data tools can read pages, fill forms, and download files. Start with a test account or read-only access, keep credentials in environment variables, and review agent actions before connecting anything that can spend money, send messages, or delete data.

Alternatives to Puppeteer

Playwright

Choose Playwright when you need WebKit as well as Chromium and Firefox, plus a test runner and an official MCP package.

Tradeoff: A larger surface to learn. Test tooling comes along even if you only need scripts.

Browser Use

Choose Browser Use when a model should decide the browsing steps from a plain task description.

Tradeoff: Less deterministic than direct Puppeteer code, and model calls add cost and variance.

Stagehand

Choose Stagehand when you want AI-assisted act and extract calls rather than hand-written selectors.

Tradeoff: You depend on a model for self-healing actions, and the API differs from classic Puppeteer scripts.

Related tools

Playwright

Browser Automation

Browser Use

Browser Automation

Stagehand

Browser Automation

Common questions

What does Puppeteer do for an AI agent?

A JavaScript library that controls Chrome and Firefox through a high-level API for automation, scraping, and testing.

Is Puppeteer open source?

Yes. This profile records the license as Apache-2.0 from the official repository.

Does Puppeteer support MCP?

Via a separate project. The Puppeteer README directs readers to chrome-devtools-mcp, a Puppeteer-based MCP server. AgentsUse has not published a separate MCP page for it yet.

Sources and freshness

Last checked 2026-10-03. Verification label: source verified, which means public claims trace to the sources above. It does not mean AgentsUse ran the tool. Spotted an error? Send a correction. Back to the tools directory.