Web scraping MCP servers

A server here extracts content or structured data from web pages.

Not here: web search.

1,765servers
1,112 remote653 localTagged by Odel's AI, which can make mistakes

Best fit

1,107 made for this, ranked by starsRemoteLocal
  1. Scavio @scavio-ai/scavio-mcpStructured web data from 36 platforms: Google, YouTube, Amazon, Walmart, Reddit, TikTok, LinkedIn8% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.8 GitHub stars.
  2. Web Search @hec-ovi/web-searchKeyless multi-engine web search, fetch, and clean-Markdown read for AI agents. No API keys.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.8 GitHub stars.
  3. Web Fetch @kylinmountain/web-fetch-mcpMCP server for web content fetching, summarizing, comparing, and extracting information8% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.8 GitHub stars.
  4. EasyBits @easybits/easybitsPersistent Linux microVMs for agents: root, internet, sub-second resume and a public URL.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.7 GitHub stars.
  5. EnConvert @enconvert/mcpUnified web ingestion for AI agents: perceive, discover, search, extract, ingest, watch, convert.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.7 GitHub stars.
  6. E2LLM @e2llm/element-to-llmE2LLM gives your AI eyes and hands in a real browser: structured perception (SiFR) plus action.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.7 GitHub stars.
  7. vibatchium @trueoriginlabs/vibatchiumStealth browser automation for AI agents — real Chrome, N parallel persistent logged-in sessions.10% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.7 GitHub stars.
  8. Reddit @mypracticaltools/reddit-mcp-serverReddit MCP Server via Apify7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.7 GitHub stars.
  9. Defuddle Fetch @domdomegg/defuddle-fetch-mcp-serverFetch content from the web, cleaned up nicely for LLMs (using Defuddle).4% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.7 GitHub stars.
  10. Camoufox @baixianger/camoufox-mcpAnti-detection browser automation with Camoufox - stealth Firefox for web scraping21% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.7 GitHub stars.
  11. Unblink @christopherdavenport/unblinkPure-Go browser for AI: fetch pages, render JS without Chromium, return clean Markdown over MCP.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.7 GitHub stars.
  12. chrome-mcp @mehmoodqureshi/chrome-mcpDrive your real Chrome over MCP: real logins and cookies, multi-tab automation, deny-all by default.9% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  13. Hilan @udah1/hilan-mcpPull Hilan (Hilanet / חילן, חילנט) payslips and Form 106 into any AI assistant.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  14. nodriver — undetected browser automation @andresolbach/nodriver-mcp-serverUndetected anti-bot Chrome automation. 70 tools, parallel isolated browsers.29% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.6 GitHub stars.
  15. Pixelcheck @xcodethink/pixelcheckReal eyes and hands on the web for AI agents. MCP-first. Local-first. Vendor-agnostic.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  16. Selenium @pshivapr/selenium-mcpSelenium Tools for MCP5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  17. Scraper @ofershap/scraperWeb scraping MCP — extract clean markdown, links, and metadata from any URL.6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.6 GitHub stars.
  18. Web Search @keenable/web-searchLive web search and clean-markdown page fetch over the Keenable web index.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.6 GitHub stars.
  19. Rendex: Rendering API for Images, PDFs & Content Extraction @copperline-labs/rendex-mcpRender HTML, Markdown, or URLs to images, PDF, or branded artifacts; extract and watch pages.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.6 GitHub stars.
  20. Trust Oracle @new1direction/trust-oracleIndependent A-F trust grade for any MCP server, watched for drift. Free, never for sale.12% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.6 GitHub stars.
  21. Amazon @pangolinfo/amazon-mcpAmazon data MCP: products, reviews, niches, WIPO design/IP + US litigation, AI search & trends.8% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from a package registry.6 GitHub stars.
  22. Fedlex Connector @fedlex-connector/fedlex-connectorSearch Swiss federal legislation: laws, articles, amendments via the Fedlex SPARQL endpoint.2% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.6 GitHub stars.
  23. Web scraping @superagnt/web-scrapingWeb scraping for AI agents: scrape, search, crawl, map any website to markdown + JSON. No browser.16% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.5 GitHub stars.
  24. Codearia Sieve @antong87/codearia-sieveWeb pages into decision-ready state: dates, numbers with units, budgeted chunks. No model, no key.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.5 GitHub stars.