Playwright MCP review

Q: How do I install Playwright MCP?

Run claude mcp add playwright npx @playwright/mcp@latest. It needs Node.js >=18. The install section above has the other options from the project's README.

Playwright MCP is a Model Context Protocol server from Microsoft that lets LLMs automate browsers through Playwright's structured accessibility snapshots.

Updated Nov 1, 2026. Review written Oct 25, 2026 from the project's own sources; usage data refreshed every Monday.

TL;DR

Playwright MCP is an MCP server that gives LLMs browser automation through Playwright. It works on accessibility snapshots, so no screenshots or vision models are needed. It suits agentic loops that need persistent browser state, while the README points high-throughput coding agents to Playwright CLI with SKILLs instead.

38.9kGitHub stars
+335new stars, week of Oct 25
~8.4Mdownloads a week, −5% vs 4-week avg
#41in MCP and agent tools this week

Playwright MCP at a glance

Category: MCP and agent tools
Maintained by: Microsoft (organization, @microsoft on GitHub)
License: Apache 2.0
Language: TypeScript, JavaScript, Shell
Repository: github.com/microsoft/playwright-mcp
Website: www.npmjs.com/package/@playwright/mcp
npm package: @playwright/mcp 0.0.83
Latest release: v0.0.83 on GitHub, Sep 28, 2026
First published: Mar 21, 2025

What Playwright MCP does

Playwright MCP lets LLMs interact with web pages through structured accessibility snapshots instead of screenshots or pixel-based input.
Playwright MCP can drive Chrome, Firefox, WebKit or Microsoft Edge, in headed mode by default or headless.
Playwright MCP generates code in TypeScript, Python, Java or C#.
Playwright MCP's browser_find tool searches the page's accessibility snapshot for text or a regex, which costs less than capturing the full snapshot.
Playwright MCP can record actions a user performs manually and return them as Playwright code, opt-in via --caps=devtools.
Playwright MCP can emulate devices such as "iPhone 15" or a generic mobile device to reduce page weight and tokens.
Playwright MCP can list and call tools a page registers through the experimental WebMCP API.

How it works

Playwright MCP runs as an MCP server launched by your client with npx @playwright/mcp@latest. Clients include VS Code, Cursor, Windsurf, Claude Desktop, Claude Code, Codex and others. Options go in the args list or environment variables. Playwright MCP can launch a browser, connect to a CDP endpoint, or attach to a running Edge or Chrome via the Playwright Extension.

How to install and start using Playwright MCP

Claude Code

claude mcp add playwright npx @playwright/mcp@latest

Codex CLI

codex mcp add playwright npx "@playwright/mcp@latest"

VS Code CLI

code --add-mcp '{"name":"playwright","command":"npx","args":["@playwright/mcp@latest"]}'

Requirements

Node.js 18 or newer
An MCP client such as VS Code, Cursor, Windsurf, Claude Desktop, Goose, Grok or Junie
The Playwright Extension, only when connecting to a running Edge or Chrome with --extension

First steps

Install Node.js 18 or newer and pick an MCP client such as VS Code, Cursor or Claude Desktop.
Add the standard config to your client: command "npx" with args ["@playwright/mcp@latest"].
Add options such as --headless or --browser to the args list to adjust browser behavior.

Who Playwright MCP is for

Exploratory automation and self-healing tests that benefit from persistent browser state and iterative reasoning over page structure.
Long-running autonomous workflows where continuous browser context matters more than token cost.
Developers who want LLM browser automation without vision models or screenshot-based input.

Limitations to know

Playwright MCP loads large tool schemas and accessibility trees into context; the README suggests Playwright CLI with SKILLs for token-limited coding agents.
Playwright MCP's --allowed-origins and --blocked-origins options do not serve as a security boundary and do not affect redirects.
Playwright MCP's extension mode works only with Edge or Chrome.
Playwright MCP's recording and video tools are opt-in and require --caps=devtools.

Is Playwright MCP growing? Usage data

In the week of Oct 25–31, 2026, Playwright MCP ranked #268 on the Devtoolwatch Board and #41 in MCP and agent tools, with a momentum score of 35 out of 100. Its main package had about 8.4M downloads that week, −5% against its own 4-week average.

New GitHub stars per week for the last 12 weeks, from GitHub's stargazer history (today's stargazers by the week they starred).

Show the weekly numbers

New GitHub stars per week for Playwright MCP
Week of	New stars
Aug 9, 2026	264
Aug 16, 2026	216
Aug 23, 2026	255
Aug 30, 2026	245
Sep 6, 2026	235
Sep 13, 2026	339
Sep 20, 2026	244
Sep 27, 2026	216
Oct 4, 2026	232
Oct 11, 2026	326
Oct 18, 2026	311
Oct 25, 2026	335

On Hacker News

Claude Code 2.0, 842 points and 413 comments (Sep 29, 2025)
Show HN: Ez FFmpeg – Video editing in plain English, 420 points and 198 comments (Dec 27, 2025)
Svelte 5 Released, 415 points and 246 comments (Oct 19, 2024)
Is-even-ai – Check if a number is even using the power of AI, 301 points and 159 comments (May 20, 2025)
NPM – "is-even", 160k weekly downloads, 247 points and 260 comments (Nov 16, 2021)

The Board ranks tools by four signals each week: new stars, new stars for the repo's size, added downloads and download growth beyond the market. How the ranking works.

Playwright MCP alternatives

Other MCP servers and agent tools on the Board in the week of Oct 25–31, 2026, by momentum:

MCP and agent tools ranked by momentum, the week of Oct 25–31, 2026
Rank	Tool	Stars	Downloads/wk
#1	REAREA is an MCP server and CLI that lets coding agents reverse engineer native binaries, JavaScript and Electron apps, .NET assemblies,…	415.5k	760.6k
#2	ha-mcpha-mcp is an unofficial Model Context Protocol server that lets AI assistants control and manage Home Assistant using natural language.	5.4k	118.1k
#3	TregTreg is a registry and proxy that gives AI agents one token for a catalog of paid third-party API endpoints and team-shared tools.	6.9k	6.7k
#4	Chrome DevTools MCPChrome DevTools MCP is an MCP server and CLI that lets coding agents control and inspect a live Chrome browser using Chrome DevTools.	54.6k	2.7M
#5	codebase-memory-mcpHigh-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 162…	47.9k	–
#6	open-computer-use👾 Open Computer Use – Open-Source Alternative to Codex Computer Use	2.8k	–

All MCP servers and agent tools, ranked

Recent Playwright MCP releases

v0.0.83, Sep 28, 2026
v0.0.82, Sep 18, 2026
v0.0.81, Sep 14, 2026
v0.0.80, Sep 1, 2026
v0.0.79, Aug 6, 2026

Playwright MCP FAQ

Is Playwright MCP free and open source?

Yes. Playwright MCP's source code is on GitHub (microsoft/playwright-mcp) under the Apache 2.0 license, so you can use it for free.

How do I install Playwright MCP?

Run claude mcp add playwright npx @playwright/mcp@latest. It needs Node.js >=18. The install section above has the other options from the project's README.

Is Playwright MCP still maintained?

Yes. The microsoft/playwright-mcp repository was last updated on Oct 22, 2026, and the latest release, v0.0.83 on GitHub, came out on Sep 28, 2026.

How popular is Playwright MCP?

Playwright MCP has 38.9k GitHub stars as of Nov 1, 2026. It gained 335 new stars in the week of Oct 25–31, 2026, against a 12-week average of 268 a week. Its main package had about 8.4M downloads that week, −5% against its own 4-week average. That made it #41 among MCP servers and agent tools on the Devtoolwatch Board.

Playwright MCP vs Browser Use: how do they compare?

Playwright MCP and Browser Use are both tracked by Devtoolwatch. Playwright MCP has 38.9k GitHub stars and Browser Use has 119.7k. In the week of Oct 25–31, 2026, Playwright MCP's main package had 8.4M downloads and Browser Use's had 1.5M. Browser Use review.

What are the best alternatives to Playwright MCP?

Other MCP servers and agent tools on the Devtoolwatch Board in the week of Oct 25–31, 2026, by momentum: REA (#1), ha-mcp (#2), Treg (#3), Chrome DevTools MCP (#4), codebase-memory-mcp (#5). The alternatives table above has their numbers.

What is Playwright MCP used for?

Playwright MCP is used to let LLMs automate browsers. It exposes Playwright through the Model Context Protocol and works on structured accessibility snapshots. The README cites exploratory automation, self-healing tests and long-running autonomous workflows as fitting uses.

Should I use Playwright MCP or Playwright CLI?

Playwright MCP fits agentic loops that need persistent state and rich page introspection. The README says coding agents may do better with Playwright CLI plus SKILLs, because CLI calls avoid loading large tool schemas and accessibility trees into the model context.

How do I use Playwright MCP with Claude Code?

Playwright MCP is added to Claude Code with one command: claude mcp add playwright npx @playwright/mcp@latest. Claude Desktop instead uses the standard JSON config from the README.

How do I use Playwright MCP in VS Code?

Playwright MCP installs in VS Code through the install button, the MCP install guide with the standard config, or the code --add-mcp CLI command. After installation it is available to the GitHub Copilot agent in VS Code.

Is Playwright MCP safe?

Playwright MCP restricts file system access to workspace roots by default and blocks file:// URLs unless --allow-unrestricted-file-access is set. It binds to localhost by default. Origin allow and block lists do not act as a security boundary.

Does Playwright MCP need a vision model?

Playwright MCP does not need a vision model. It operates on Playwright's accessibility tree. Vision is an optional extra capability enabled with --caps.

Sources

Playwright MCP README badge

Maintainers can show the tool's weekly Board rank in their README:

[![Devtoolwatch](https://index.devtoolwatch.com/badge/playwright-mcp.svg)](https://index.devtoolwatch.com/tools/playwright-mcp/)

How this page is made: Devtoolwatch writes each review from the project's README, releases, package metadata and website, with an AI model, then checks every claim against the source it cites and drops what the sources don't support. The numbers come from Devtoolwatch's daily GitHub snapshots, GitHub's stargazer history, and the npm and PyPI download APIs. Spotted an error? Email [email protected].