Web scraping MCP servers

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

Not here: web search.

1,762servers
1,108 remote654 localTagged by Odel's AI, which can make mistakes

Best fit

1,105 made for this, ranked by starsRemoteLocal
  1. TikTok Profile Stats @mrocker/tiktok-profile-statsTikTok profiles (followers, bio) and per-video stats by handle or URL. No login. Pay per result.15% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
  2. Inferventis — Financial Data, News & Web MCP @jonathanpchapman/inferventisLive financial data MCP: FX, crypto, stocks, news, URL reader. x402 on Base: $0.001/call.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  3. Superfetch @j0hanz/superfetchIntelligent web content fetcher MCP server that converts HTML to clean, AI-readable JSONL format6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  4. Reddit Scraper Pro (Apify) – posts, comments, search, users, subreddits @harshmaur/reddit-scraper-proReddit posts, comments, search, users & subreddits as JSON via Apify. No Reddit API key. $1.50/1k12% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
  5. Link Preview API @cmondillo/link-preview-apiPay-per-call Open Graph/link preview metadata for AI agents. $0.01 USDC per call, no signup.6% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  6. webrecipe @pillsoon/webrecipeSave how to read a public web page once, then fetch fresh items as JSON without a browser.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  7. HTML to Markdown @mukundakatta/html-to-markdown-mcpConvert HTML to Markdown or strip to plain text. For web-scraping agents.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  8. Metalift @metalift-ai/metaliftMetalift MCP for AI agents: web search (2 credits), scrape, crawl, and map tools6% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  9. Snow Leo Leads & Contacts @magzhansan/snowleo-leadsLocal business leads from Google Maps, 7 directories, DuckDuckGo and OpenStreetMap, with contacts7% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
  10. Snow Leo Marketplaces & Prices @magzhansan/snowleo-marketplacesPriced listings from Amazon, Shopify stores, the App Store, Google Play, Airbnb and Redfin6% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
  11. CyberWareX Web Access @cyberpunk11147/cyberwarex-web-accessWeb access for agents: fetch, extract, screenshot, or PDF any public URL.12% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from Docker.
  12. URLpipe @urlpipe/mcpRead any page with its JavaScript run: Markdown, screenshots, metadata, console errors, Lighthouse.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  13. Shipshape Data docs @shipshapedata/shipshape-data-docsDocs Q&A: search 169 data and AI guides, fetch any page as markdown. Read-only, keyless.2% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  14. GoAI Moat Web Scrape @goaimoat/web-scrapeScrape any URL into clean LLM-ready Markdown, or crawl a whole site.13% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  15. Anthony Hunt · SHAART Agency docs @anthonyhunts/shaart-agency-docsSearch anthonyhunts.com and read any page as Markdown: GoHighLevel and AI automation guides3% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  16. MRX Public MCP @underwriter-mrx/mrx-mcpRead-only, source-linked mineral-rights education from MineralRightsXchange.com.3% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  17. Steady API Scrapers @steadyapi/scrapersYouTube transcripts, Google Hotels prices, Google Ads Transparency, Google Trends and Threads posts.8% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
  18. Evercraft Web @jgaethle10/evercraft-webRead public web content with Evercraft; search and browser automation stay behind safety gates.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  19. Bandcamp crate digging @venut-technologies/bandcamp-mcpDig Bandcamp from any AI assistant: search artists, albums, labels and tracks, browse genre tags.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  20. Straight Line Golf Pro @straightlinegolfpro/straight-line-golf-pro-mcpRead-only discovery and retrieval of current public Straight Line Golf Pro coaching resources.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  21. ViewportWitness by Apex Labs @baffles78/viewport-witnessAgent website testing: browser QA, visual regression, Markdown extraction, and security checks.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
  22. SnoopScan @snoopscan/snoopscanWeb scraping for agents: scrape, crawl, map, search and extract pages as clean markdown or JSON.11% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; you sign in with your account.
  23. Exnos @golproductions/exnosRead-only live state of your Chrome tabs for AI agents: fields, console errors, network, storage.9% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
  24. x402 Tools @foxxx009/x402-tools-mcpPay-per-call research tools for AI agents, paid via x402 USDC. Remote MCP, one free tool.9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.