
Playwright MCP servers: which one to use
Use Microsoft's Playwright MCP to let an AI browse and test sites. Pick a community server for anti-bot stealth, parallel browsers or reading test reports.
- Servers
- 58
- active
- Official
- 1
- by Playwright
- Hosted
- 6
- nothing to install
- Local
- 40
- runs on your machine
Microsoft makes the official Playwright MCP server, and it is the right start for most people. It lets Claude or Cursor open a real browser, click through pages and test your web app. Community servers cover narrower jobs, such as hiding automation from anti-bot checks, running several browsers at once, or reading your Playwright test results. Whichever you pick, the AI controls a real browser, so keep it away from accounts you cannot afford to damage.
Good for
- Developers who want an assistant to check the app they just built in a real browser
- QA engineers who write and debug Playwright tests with AI help
- Teams running several AI agents that each need their own browser
- Anyone reading a site that blocks automated browsers
Which Playwright MCP server to use
Start with Microsoft's Playwright MCP. Choose Playwright Stealth or Invisible Playwright only when a site blocks automation, Playwright Parallel for several agents at once, and Playwright Report to analyze failed tests.
| Server | Official | Runs | Best for | Anti-bot stealth | GitHub stars |
|---|---|---|---|---|---|
| Playwright MCPThis page | Official | Local | Browsing and testing | – | 38.9K |
| ExecuteAutomation | Community | Local | Test code, scraping | – | 5.7K |
| Invisible Playwright | Community | Local | Anti-bot browsing | Yes | 886 |
| Playwright Stealth | Community | Local | Scripted browsing | Yes | – |
| Playwright Parallel | Community | Local | Parallel agents | – | 19 |
| Playwright Report | Community | Local | Test results | – | 0 |
All Playwright MCP servers
58 active · 1 official · official first, then by GitHub stars
Playwright MCPOfficialLocalPlaywright MCP server
38.9K GitHub stars
mcp-playwrightLocalNo sign-inPlaywright Model Context Protocol Server - Tool to automate Browsers and APIs in Claude Desktop, Cline, Cursor IDE and More 🔌
5.7K GitHub stars
invisible_playwright_mcpLocalAI browser agent: browses, clicks, types, and reads real web pages from plain-English instructions.
886 GitHub stars
playwright-stealthLocalBrowser automation using Playwright with optional stealth mode to bypass anti-bot protection.
81 GitHub stars
infy_mcp_playwrightLocalMCP Playwright server for web app testing automation via HTTP API
61 GitHub stars
mcp-playwright-cdpLocalModel Context Protocol server for Playwright with Chrome DevTools Protocol support
52 GitHub stars- chrome-debug-mcp
An MCP server to allow you to debug webpages using LLMs
45 GitHub stars
Luvv-MCPLuvv-MCP 是一个 AI 原生的前端安全分析引擎,接入 Claude Code 后,你说一句"帮我审计这个网站",AI 就能自动完成技术栈识别(1.8 万条规则覆盖 CMS/框架/CDN/中间件)、泄露扫描(阿里云 AK、GitHub Token、Webhook、JWT 等 60+ 类敏感信息)、漏洞审计(XSS、开放重定向、CSRF、IDOR、硬编码密码 6 类自动检测)、产品指纹识别(HTTP 响应头/Body/Favicon Hash 多维匹配),还能自动发现 SourceMap 并还原混淆源码。核心价值:把 SRC 挖洞和授权渗透中"人工看 JS 源码找漏洞"的环
42 GitHub stars
Playwright-AI-Agent-POM-MCP-ServerPlaywright AI Agent POM MCP ServerPlaywright AI Agent using Page Object Model (POM) architecture with MCP Server integration for automated web and mobile testing
39 GitHub stars
@roxybrowser/playwrightLocalMCP server for RoxyBrowser automation - manage browser instances and get CDP endpoints
38 GitHub stars
playwright-mcpLocalPlaywright integration for ModelContext
32 GitHub stars- @ohos-ports/playwright-mcpLocal
Playwright integration for ModelContext
32 GitHub stars - @iflow-mcp/qabyai-playwright-mcpLocal
Playwright integration for ModelContext
32 GitHub stars - playwright-parallel-mcpLocal
MCP server for parallel browser automation with Playwright
19 GitHub stars
Playwright Cloud BrowsersHostedManaged Playwright cloud browsers for scalable browser automation and AI agent workflows
15 GitHub stars- playwright-autopilot
Autonomous Playwright E2E test fixer — Claude Code plugin with action capture, DOM snapshots, page object scanning, and business flow memory
12 GitHub stars - mcp-playwright
Playwright MCP Server 一个基于 FastMCP 框架的专业级 Playwright 浏览器自动化 MCP 服务器,为 LLM 应用提供强大的网页交互和数据提取能力。
10 GitHub stars - auto-captcha-solver
Universal captcha solver for AI Agents using Playwright — hCaptcha + reCAPTCHA v2 via NopeCHA API. Python lib, CLI, MCP server, Hermes skill.
9 GitHub stars - claude-playwright
Seamless Claude Code ↔ Playwright integration with intelligent caching. Transform browser automation with AI-aware selector resolution and persistent sessions.
9 GitHub stars - playwright-wizard-mcpLocal
MCP server providing Playwright test generation wizard with intelligent prompts and best practices
7 GitHub stars
Show all 58Show fewer
- playwright-sse-mcp-serverLocal
playwright/mcpをsseで外部から触るためのやつ
7 GitHub stars - playwright-record-mcp
playwright-mcp with video record
7 GitHub stars - playwright-trace-mcpLocal
Playwright MCP with trace viewer and video recording capabilities
6 GitHub stars - playwright-mcp-for-operit
Playwright MCP for Operit - 在 Operit(Android/proot) 中运行官方 @playwright/mcp 网页自动化插件,headless Chromium + 24 个 browser_* 工具,替代失效的内置 browser 包
6 GitHub stars
playwright-mcp-advancedLocalAdvanced Playwright Tools for MCP
5 GitHub stars- playwright-mcp-worldLocal
Community-driven Playwright Tools for MCP with additional features
5 GitHub stars - Playwright-search-MCP
Free local Google Search MCP server — Playwright-driven, persistent Chrome profile, headless-by-default with CAPTCHA reveal
5 GitHub stars - ultimate-playwright-mcpLocal
Multi-agent Playwright MCP server with tab group isolation — multiple AI agents share one Chrome browser
3 GitHub stars - playwright_for_android
Playwright for Android app development — an MCP server that drives an Android emulator from your LLM: build, install, launch, and automate app features end-to-end with a unified logcat-and-actions timeline.
3 GitHub stars - playwright-react-debug-mcpLocal
MCP server for frontend debugging with Playwright
2 GitHub stars - Playwright_MCP_Server
This repository documents hands-on experience working with Playwright Test Agents through MCP.
2 GitHub stars
playwright-trace-decoder-mcpLocalMCP server for unpacking and analyzing Playwright trace.zip archives
1 GitHub stars- playwright-mcp-multitab-enhancedLocal
MCP server for automated job applications with multi-tab browser control
1 GitHub stars
playwright-network-chaos-mcpLocalMCP server that gives AI agents dynamic network chaos control over Playwright browser sessions
1 GitHub stars
Playwright E2E MCPHosted or LocalDiagnose and fix Playwright tests: CI reports, traces, verified fixes, flaky verdicts, audits.
0 GitHub stars
playwright-report-mcpLocalRun Playwright tests and surface structured results for AI agents doing test failure analysis.
0 GitHub stars
TurnSignal for PlaywrightLocalRead-only Playwright CI results for agents: new vs known failures, flaky tests, test history.
0 GitHub stars
playwright-mcp-wslLocalDrop-in WSL-safe wrapper for @playwright/mcp. Prevents GPU crashes on Windows Subsystem for Linux by disabling Chromium GPU acceleration through the dxg bridge.
0 GitHub stars
playwright-spatial-layout-mcpLocalMCP server that gives AI agents geometric spatial awareness of web page layouts using Playwright
0 GitHub stars- playwright-mcp-automationLocal
Lightweight Playwright MCP automation - zero config, works with all AI IDEs
0 GitHub stars - playwrightmcpserverLocal
MCP Playwright Test Generator - Generate Playwright tests from JSON specifications
0 GitHub stars - agentic-playwright-mcpLocal
Fork of official Playwright MCP with multi-agent tab isolation via targetId
0 GitHub stars - benjaminkitt-nix-playwright-mcpLocal
Playwright MCP Server - A Model Context Protocol server for browser automation with Chrome and Firefox support
0 GitHub stars - playwright-parallel-mcpLocal
MCP server for parallel browser automation with isolated Playwright sessions
0 GitHub stars - playwright-mcp-hostsLocal
Playwright MCP with custom hosts mapping, like SwitchHosts for browser automation
0 GitHub stars - vscode-playwright-mcpLocal
Playwright for VS Code. MCP server that gives LLMs full mouse, keyboard, and visual control over VS Code via Playwright's Electron API.
0 GitHub stars - ohos-playwright-mcpLocal
MCP server for HarmonyOS / OpenHarmony ArkWeb — playwright-core over CDP. 61 tools: navigation, input, network, cookies, storage, snapshot, tracing. The ohos counterpart of @playwright/mcp.
0 GitHub stars - playwright-mcpLocal
Enhanced Playwright MCP with intelligent browser console message filtering
0 GitHub stars - playwright-mcp-tabbedLocal
Playwright MCP server with per-tool tab_index support for parallel agent use
0 GitHub stars - xtr-dev-mcp-playwright-webx11Local
Playwright MCP Server with noVNC Display - A Docker Compose setup that runs the Microsoft Playwright MCP server with a virtual X11 display accessible via noVNC in your web browser
0 GitHub stars - mcp-agentsLocal
A Model Context Protocol (MCP) server that uses Playwright to fetch and render web pages to Markdown or JSON.
0 GitHub stars - Playwright URL CheckHosted
Check whether a URL can be opened. No browser is launched.
0 GitHub stars - playwright-mcpLocal
Playwright Tools for MCP
0 GitHub stars - playwright-selector-guard-mcpHosted
Playwright selector risk checks with healing suggestions and receipts.
0 GitHub stars - playwright-slim
playwright MCP (73% less tokens). Quick setup: npx playwright-slim --setup
0 GitHub stars - PlaywrightGenHosted
Approved test intent, reviewed Playwright automation and run evidence, inside your editor.
0 GitHub stars - playwright-mcp-serverLocal
MCP server for generating Playwright tests
1.1K/wk downloads a week - playwright-mcp-serverHosted
Headless browser primitives for AI agents when sites need real JS rendering.
Questions
Is there an official Playwright MCP server?
Yes. Microsoft publishes Playwright MCP, and it is free and open source under Apache-2.0. It needs no account or API key.
Which Playwright MCP server should I use?
Microsoft's, unless you have a specific need. It supports Chrome, Firefox, WebKit and Edge, and it has the most GitHub stars by a wide margin.
Can it use my logged-in accounts?
With Microsoft's server, yes. The default profile keeps logins between sessions, and extension mode connects to your open Chrome or Edge. Use the isolated option for throwaway sessions.
Is Playwright MCP safe to run?
Microsoft says it is not a security boundary. Update to 0.0.40 or later, which fixes a flaw where a website could reach a locally running server, and keep approval on for the tool that runs code.
Which one gets past anti-bot checks?
Playwright Stealth and Invisible Playwright. The stealth server has an optional stealth mode, and Invisible Playwright describes itself as invisible to anti-bots. Check a site's terms before you automate it.
Can several AI agents use it at once?
Not well with Microsoft's server and the same saved profile, which gives 'browser already in use' errors. The Playwright Parallel README says it starts a separate browser for each session to avoid this.
How is ExecuteAutomation different from Microsoft's?
It is a community server that adds test code generation, scraping and device emulation. Its README lists 143 device presets, for example iPhone 13 and iPad Pro 11.
Why does Playwright MCP use so many tokens?
Each action returns a snapshot of the page, and the tool list also loads into context. Microsoft rates its CLI as lighter for coding agents.
Can it run without a visible browser window?
Yes. Add the headless option to the server arguments with Microsoft's server. The browser is visible by default, and its Docker image only runs headless Chromium.