Browser Automation MCP servers

Official Microsoft Playwright MCP server, enabling LLMs to interact with web pages through structured accessibility snapshots

6 tools · Connect to an existing browser, Control session persistence

Browser automation for web scraping and interaction

2 tools · navigate, screenshot

Automate browser interactions in the cloud (e.g. web navigation, data extraction, form filling, and more)

6 tools · act, end

Official Firecrawl server with live browser interaction (`firecrawl_interact`) — navigate, click, type, and scroll on a page before extraction, for JS-heavy or auth-gated sites. Also scrape, crawl, map, search, and extract.

6 tools · firecrawl_agent, firecrawl_crawl

Capture webpages as images or PDFs with full-page and element targeting, device and viewport controls, location selection, page interactions, blocking options, and configurable output.

6 tools · cache, capture_screenshot

Lightweight browser MCP server for LLM agents. Runs JavaScript, follows links, fills forms, manages cookies, and returns low-token BlockMaps from a single native binary without Chrome.

Token-efficient browser automation for LLM agents: semantic page snapshots and stable element IDs instead of raw DOM or screenshots, driving Chrome over CDP.

An MCP server in Go for Lightpanda, the ultra fast headless browser designed for web automation

Local-first MCP server (@peekdev/mcp) + Chrome extension that records your real authenticated browser session (rrweb DOM + console + network, PII-masked, to local SQLite) and lets AI coding agents query it or drive the browser. No cloud, no telemetry.

🦀 Lightweight browser automation MCP server in Rust with zero dependencies.

Web scraping, structured data extraction, and screenshots with JavaScript rendering, residential proxy rotation, and automatic retries.

Free webpage screenshot capture API with format, viewport, and dark mode options. No API key required.

An MCP Server Integration with Apple Shortcuts

MCP server that compiles AI browser automation into deterministic `.tap.json` plans (25-op closed union, zero runtime LLM), runs on your logged-in Chrome, and detects drift via semantic fingerprint diff when sites change. 65+ open community taps on 40+ sites.

Zen Browser automation via WebDriver BiDi. 20 tools for navigation, form filling, screenshots, and JavaScript evaluation. No Selenium or Playwright required.

An MCP server using Playwright for browser automation and webscrapping

Access and organize your open browser tabs, bookmarks, and recently opened links across every macOS browser from your AI client. Powered by the free SupaSidebar app.

In-browser debug bridge that injects into your running web app, so an agent can read the DOM, console logs and network requests, and click/fill/inspect the page. Runs inside the page (not an external CDP driver), so it works in any browser and inside Cypress/Playwright runs.

Point your coding agent at a URL and get a real-browser QA audit: broken signup/login/checkout flows, JS console errors, missing analytics, consent + security headers, mobile tap targets, and accessibility — returned as machine-verified findings graded A–F. Free 60-second audit, no signup.

Python MCP server for Selenium WebDriver — 84 tools for browser automation, element interactions, assertions, and self-healing locators, plus codegen for Java TestNG / JUnit 5 / Cucumber / pytest / C# NUnit / Playwright and CI pipelines (GitHub Actions / Jenkins / GitLab CI). No ChromeDriver setup needed.

AI-native web scraper MCP server. Single binary, returns clean markdown, MIT-licensed Firecrawl alternative.

Multi-tier fallback chain for fetching web content as clean markdown. Handles tweets, YouTube, arXiv, PDFs, and regular pages with 9 fallback strategies.

Self-hosted stealth browser automation for unattended agents. Drives real consumer Chrome via Patchright, holding N parallel persistent logged-in profiles on one daemon so sessions survive across runs. Encrypted credential vault with TOTP and IMAP email-code 2FA, a renderer-free TLS-impersonating fetch lane that reuses session cookies, token-frugal page extraction, and prompt-injection scanning on by default.

An MCP server for interacting with Apple Reminders on macOS

Design-debugging context for AI agents from a real Chrome: which CSS rule WINS with file:line, blast radius (how many other elements that rule styles) + a scoped-fix selector, WordPress/Elementor origin attribution, interaction timelines, live CSS fix trials, and pixel/alignment/WCAG-contrast audits. No AI inside — deterministic CDP.

The solve brain for CAPTCHA workflows: identifies CAPTCHA type and task parameters, serves CapMonster Cloud docs, creates solving tasks, and retrieves results via the CapMonster Cloud API. In the official MCP Registry as `io.github.CapMonsterCloud/capmonster-mcp-captcha-solver`. `uvx capmonster-mcp` or `npx capmonster-mcp`

Precision visual captures for AI agents — real-colors terminal, syntax-highlighted code, browser, markdown, HTML, and git-diff as PNG/JPEG/PDF/GIF, plus multi-shot sequences and doc embedding. Rendered fully local via Playwright (SSRF-safe by default).

Connect to Perplexity Comet browser for agentic web browsing, deep research, and real-time task monitoring.

Pay-per-call web scraping for AI agents via x402 micropayments on Base. Six tools: scrape, extract structured data, screenshot, metadata, browser session, workflow. No signup, no API keys — just USDC. HTTP + MCP transports.

A `minimal` server/client MCP implementation using Azure OpenAI and Playwright.

Web scraping MCP server for AI agents. 6 tools: clean content extraction, structured scraping with CSS selectors, full-page screenshots via Playwright, link extraction, metadata extraction (OG/Twitter cards), and Google search. Free tier with x402 micropayments.

Drive a local Playwright browser through residential proxies with the agent choosing the exit country/city and holding one sticky identity per task. 10 tools: launch, navigate, accessibility snapshot, click, type, content extraction, exit-IP verification, and per-task identity rotation. Free trial, no card.

Rent real residential Chrome browsers from real humans — per-minute billing ($0.01/min), real fingerprints, residential IPs, no datacenter signals. MCP-native, with SDKs for Python, JS, LangChain, CrewAI, n8n. Crypto payouts.

Glint Render (by Moyu) — web rendering & evidence API for AI agents: screenshots, PDFs, OG images, monitoring, extraction and tamper-evident evidence. Free tier included. `npx -y ottersnap-mcp`

A FastMCP-based tool that fetches Bilibili's trending videos and exposes them via a standard MCP interface.

Production HTTP 402 micropayment web scraper, Llama-3 digest, and Twitter intelligence engine for AI agents on Base L2. Zero API keys, pay-per-call in USDC.

Compresses ~100K-token HTML into 2-5K-token structured maps while preserving every actionable element. AI agents can read and interact with any web page at 97% fewer tokens.

Zero-config diagnostics for MCP servers. Auto-discovers configs across Claude Code, Cursor, VS Code, Windsurf, and Claude Desktop, then tests connections via JSON-RPC handshake, audits security issues, and benchmarks latency. Also runs as an MCP server itself.

A Model Context Protocol (MCP) integration that provides Claude Desktop with autonomous browser automation capabilities.

Extract structured data from any website. Just prompt and get JSON.

An MCP server paired with a browser extension that enables LLM clients to control the user's browser (Firefox).

browser-use packaged as an MCP server with SSE transport. includes a dockerfile to run chromium in docker + a vnc server.

Intent-aware visual verification for coding agents: the agent declares what a UI change should affect, and SnapDiff diffs the page against a baseline and flags anything that changed outside that intent for review or rollback — local screenshot capture via Playwright.

Web scraping with four escalating fetch tiers (httpx, TLS impersonation, Playwright, stealth) that only escalate on a detected block. Returns clean markdown, or saves large scrapes into SQLite tables you query with read-only SQL to keep them out of the context window. 24 tools covering crawls, products, contacts, PDFs and change tracking. MIT, no API key.

Web browsing superpowers for AI agents: 13 MCP tools (scrape, extract, crawl, map, search, academic papers, batch, research, monitor, session). 3-tier auto-fallback ladder with free Cloudflare bypass, no API keys. Self-hosted Firecrawl alternative.

Drive the real, logged-in Chrome you already have open — cross-platform, zero dependencies beyond MCP SDK, token-budgeted reading (90%+ savings), complex SPA forms and matrix surveys, and 24/7 headless Ubuntu support.

iOS Safari debugging via ios-webkit-debug-proxy — MCP server with full WebKit Inspector Protocol support (DOM, CSS, Network, Storage, Debugger, and more)

A Model Context Protocol server providing web automation capabilities through Selenium WebDriver

Playwright MCP-compatible browser automation using CloakBrowser Chromium, available as an npm package and Docker image.

Screenshot, PDF, and HTML rendering API for AI agents. Capture any URL or raw HTML as PNG/JPEG/WebP/PDF with batch processing, geo-targeting, async webhooks, and MCP-native integration. Free tier included.

An MCP Server enables AI assistants to autonomously test, debug, and validate web applications.

Headless browser automation and web scraping infrastructure. Exposes the Browserless smart scraper API to LLM clients over MCP.

Web intelligence API for AI agents — screenshot capture, PDF generation, page metadata extraction, and 75+ specialized data extractors for news, social media, SERP, pricing, and more. Free tier included.

AI virtual staging for real estate. Stage empty rooms, beautify floor plans into 3D renders, classify room images, generate property descriptions, and get style recommendations.

MCP server exposing 220 browser automation tools across 13 capability tiers. Chrome + Firefox via CDP + BiDi. No Node.js, no Chromium download — uses your existing browser. Stealth mode, Lighthouse audits, multi-action YAML batching, raw CDP/BiDi access, structured errors with LLM-actionable suggestions.

An MCP python server using Playwright for browser automation,more suitable for llm

Pipe one-click browser-tab screenshots into any MCP client, 100% local — a companion extension captures, the local server stacks them and serves them to your LLM on demand.

MCP server adapter for the lightweight Rust headless browser Obscura — high-performance web scraping with anti-detection. Perfect for AI agent automation. Server can run [locally](https://github.com/Metadrama/obscura-mcp) or as [hosted endpoint](https://glama.ai/mcp/connectors).

Token-budgeted web fetch for AI agents. Auto-routes between Trafilatura, Jina Reader, FireCrawl, and pypdf based on URL pattern. `estimate_tokens` before `fetch_url`, 6h Redis cache, server-side `max_tokens` truncation. Open source MCP server (MIT) plus hosted REST API at [agentfetch.dev](https://www.agentfetch.dev) — 500 free fetches/mo, no card.

Self-hosted stealth browser for AI agents: own Firefox engine fork (C++-level fingerprint spoofing) + Rust MCP runtime. page_a11y semantic vision pierces shadow DOM, login_state detection, evidence recording, Android personas. 19 tools, 6 platforms. Glama 100% quality score.

MCP server + Chrome extension that gives AI agents control of the user's real browser with existing sessions, logins, and cookies. No headless browser, no re-authentication.

An MCP server that enables free web searching using Google search results, with no API keys required.

MCP server + Chrome extension that connects Claude Code to Chromium-based browsers (Arc, Vivaldi, Brave) where the official Claude in Chrome extension doesn't work — navigate, click, read pages, run JS, and record GIFs in the user's real browser.

Code-aware browser testing for coding agents. Thirteen tools read routes, forms, field names, and existing tests before running Playwright scenarios, retaining regression history and producing HTML reports with screenshots. No internal LLM calls or API key. `npx -y vibe-testing@latest --mcp`

Security-hardened Chrome automation with post-quantum encryption (ML-KEM-768 + ChaCha20-Poly1305), secure credential vault, memory scrubbing, and audit logging. 22 tools for browser automation and secure logins.

Browser automation capabilities using Puppeteer, both support local and remote browser connection.

Fetch content from login-protected web pages (Notion, Google Docs, Jira, Confluence, etc.) by opening a real browser for authentication with persistent session caching.

Website screenshot API for AI agents. Real Chromium rendering, full-page capture up to 20,000px, country/geo proxies, and automatic ad & cookie-banner removal (cleaner images, fewer vision tokens). Two tools: `capture_screenshot`, `capture_full_page`. Install: `npx -y site-shot-mcp`.

A Python-based MCP server for browser automation, testing, and bypassing bot-detection using SeleniumBase CDP Mode with a Chromium browser. The `solve_captcha` tool can handle CAPTCHAs that expect a click.

MCP server bridging Claude Desktop with Selenix for browser automation and testing. Create, run, debug, and manage browser tests through natural language.

Safari Web Extension + Node.js MCP bridge giving Claude Desktop full control over Safari — navigate, read pages, click elements, fill forms, and manage tabs. No Playwright or WebDriver dependency.

Fetch any web page's HTML (or AI-parsed JSON, or Google results) through the ScrapeUnblocker anti-bot API (Cloudflare, DataDome, PerimeterX, Akamai, Shape), using your own API key. Three tools: `fetch_html`, `fetch_parsed`, `google_search`. Install: `npx -y scrapeunblocker-mcp`.

URL to screenshot, PDF, or structured extract for AI agents via MCP and REST.

Browser and mobile app automation using WebdriverIO, enabling AI agents to control browsers, interact with web elements, and automate native Android and iOS apps via the WebDriver and Appium protocols.

Firefox browser automation via WebDriver BiDi for testing, scraping, and browser control. Supports snapshot/UID-based interactions, network monitoring, console capture, and screenshots.

An MCP Server that enables AI assistants to interact with your local browsers.

Fetch YouTube subtitles and transcripts for AI analysis

Local-first web reading with a tier ladder that escalates from plain HTTP to a real browser when a page pushes back. Pages as clean Markdown, whole-site crawls, keyless search, repeated page structure induced into rows, and a REST API. Attempts supported captchas on your own machine with optional local ONNX models, and parks the rest in a human dashboard. Labels blocked and low-quality pages instead of handing back a challenge screen as content. No API keys, no paid solvers, no telemetry.

A MCP server that supports searching for Bilibili content. Provides LangChain integration examples and test scripts.

A fully functional MCP server and CLI for YouTube to automate YouTube operation

Screenshot API for AI agents — capture any website as PNG, JPEG, WebP, or PDF with device emulation, dark mode, ad blocking, and cookie banner removal. Free tier included.

An MCP server for browser automation using Playwright

Automate your local Chrome browser

Native Safari browser automation for AI agents with 80+ tools. No Chrome dependency, optimized for Apple Silicon with 60% less CPU overhead.

Read any URL as a clean AIDocument (Markdown + structure) through Lyrenth's cached index.

MCP server for screenshots, PDFs, OG images, and narrated video recording from Claude Desktop, Cursor, and Windsurf.

Transcriptor MCP is your choice when you need transcripts and metadata for AI, summarization, or content analysis

Chrome automation over raw CDP — no Playwright, no extension bridge, no single-tab limit. Accessibility-tree refs cached across calls survive rerenders, so a page snapshot costs 1,124 chars against Playwright MCP's 6,084. Server-side plan executor runs multi-step flows in one tool call. Ships a Python API for LLM-free scripting. MIT.

Record what you do on any website and securely automate it forever. Replays browser actions in headless Playwright with encrypted credentials and AI-powered selector healing.

An MCP Selenium Server for controlling browsers using natural language in Cursor IDE. Perfect for testing, automation, and multi-user scenarios.

Open-source MCP-native browser agent with human takeover via noVNC, reusable auth profiles, and approval/audit rails. Playwright + FastAPI, Docker-based isolated sessions, stdio bridge for Claude Desktop and Cursor.

Autonomous LLM-powered web crawler with 17 browser tools and goal-driven agent. Single binary, stealth browsing, 25 LLM providers.

An MCP Server that autonomously debugs web applications with browser-use browser agents

Real-time web intelligence with freshness timestamps. GitHub, HN, Scholar, arXiv, YC, jobs, finance, package trends — every result stamped with how old it is.

Website & web-app testing built for AI agents rather than raw browser bindings: 66 Playwright tools with hard assertions, auto form-fill, persistent authenticated sessions (incl. interactive login for 2FA/SSO), network mocking, real INP, and accessibility/SEO/GEO (llms.txt, AI-crawler access, WebMCP) + Lighthouse audits.

Chrome extension + MCP server that gives AI agents control of your real, logged-in Chrome browser (not headless). 34 tools, up to 10 concurrent color-coded sessions. Works with Claude Code, Cursor, VS Code, Codex CLI, and ZCode.

Browse the web using filesystem commands (ls, cd, grep, click). 38 MCP tools map Chrome's Accessibility Tree to a virtual filesystem via a Chrome Extension.

MCP server + Chrome extension that drive the user's real, logged-in Chrome over a local WebSocket. 59 token-efficient web-dev tools: compact element refs instead of screenshots, server-side table filtering, visual regression, accessibility/SEO/security audits, network mocking. Also runs on ChromeOS/Crostini.

Drives a browser (or a native Expo/React Native app on iOS/Android) and your dev server onto one correlated timeline, so a browser console error and the backend stack trace from the same moment line up. Includes `repro` action sequences, a unified log/network query API, and an Electron cockpit.