MCP server category
Web search and extraction MCP servers: comparison and full list
The Web search and extraction category lists 159 open-source MCP servers, checked on GitHub. The best scored on our grid are Exa MCP server, Brave search MCP server, and Arxiv MCP server. 71% received a commit in the last three months.
159servers
71 %active (3 months)
33official
27stars (median)
What Web search and extraction MCP servers are for
Web search and extraction MCP servers help an AI agent find pages and retrieve their contents. They support research teams, information specialists and people who need to compare online sources. Depending on the server, you can run searches, crawl a site or extract structured information. Exa MCP server describes web search through its API. Trieve provides functions including crawling, search and information retrieval within datasets. A practical request might be: “Find relevant pages about this topic and compare their claims while keeping the source links.” The usefulness of the answer depends on what the server can actually retrieve and preserve.
Example uses:
- Find relevant pages through Exa before preparing a comparison of online source material.
- Extract content from accessible pages to assemble a research file for ongoing monitoring.
- Collect and search information within datasets using the documented functions available through Trieve.
- Find arXiv papers and retrieve their bibliographic metadata through Arxiv MCP server.
The category in figures
Of the 159 servers in the category (5% of the directory), 33 are official vendor implementations, 91% declare a licence, 41 offer a remote server and 43 need at least one API key or token according to the MCP registry. The median repository has 27 stars.
All Web search and extraction MCP servers
Click a header to sort. Score out of 100: adoption, maintenance and framework (licence, registry, official status); risk: system access, secrets, licence, abandonment.
| Server | Stars | Last commit | Licence | Risk | Score |
|---|---|---|---|---|---|
| Wigolo Local web search, fetching, crawling, extraction and research, with multi-engine search, local ML reranking, a SQLite cache and JavaScript r | 5,499 | Oct 24, 2026 | — | high | 77 |
| Exa MCP server ◆ Real-time web search using the Exa AI Search API. | 5,163 | Oct 19, 2026 | MIT | low | 97 |
| JobSpy Search LinkedIn, Indeed, ZipRecruiter, Glassdoor and regional job boards | 4,552 | Oct 23, 2026 | MIT | low | 82 |
| Tolgee Your app's translations in Tolgee: search keys, create translations, trigger machine translation | 4,139 | Oct 24, 2026 | — | moderate | 71 |
| Arxiv MCP server Search arXiv papers, read original LaTeX sections, get BibTeX from arXiv metadata and set topic watches, with papers stored locally. | 3,232 | Oct 8, 2026 | Apache-2.0 | low | 85 |
| Trieve ◆ Crawl, embed, chunk, search and retrieve information from datasets through Trieve. | 2,721 | Jan 25, 2026 | MIT | low | 69 |
| Tavily MCP MCP server for advanced web search using Tavily | 2,454 | Oct 22, 2026 | MIT | moderate | 79 |
| Open webSearch Multi-engine web search across Bing, Baidu, DuckDuckGo, Brave, Exa and CSDN. | 1,874 | Oct 8, 2026 | Apache-2.0 | low | 78 |
| Duckduckgo MCP server Web search using DuckDuckGo, with content fetching and parsing. | 1,550 | Sep 4, 2026 | MIT | low | 72 |
| Brave search MCP server Web search using Brave's Search API. | 1,525 | Oct 22, 2026 | MIT | moderate | 88 |
| AutoTS Automated time series forecasting with model search, anomaly detection, and event risk analysis | 1,434 | Oct 15, 2026 | MIT | low | 77 |
| SearXNG Search Web search via SearXNG. | 1,329 | Oct 22, 2026 | MIT | moderate | 81 |
| Fetcher MCP Fetch web page content using a Playwright headless browser, supporting JavaScript rendering, content extraction and Markdown or HTML output. | 1,087 | Jan 14, 2026 | MIT | low | 55 |
| Kagimcp Search the web with Kagi Search. | 547 | Jul 7, 2026 | MIT | low | 60 |
| G search MCP Google search with parallel queries for multiple keywords. | 273 | Jun 14, 2025 | MIT | moderate | 41 |
| Nimrod Web search with quality scores; web pages are read on your computer by the local server. | 258 | Oct 3, 2026 | — | high | 65 |
| AffectUI Search, read and install agent-ready React components, screens and blocks for AI products. | 238 | Oct 24, 2026 | MIT | low | 69 |
| Unsplash MCP server Unsplash image search. | 237 | Oct 12, 2026 | MIT | low | 69 |
| MinerU Open MCP ◆ Convert PDFs, Word and PowerPoint files, images and spreadsheets to Markdown via the MinerU document parsing API. | 229 | Sep 22, 2026 | Apache-2.0 | moderate | 79 |
| MCP server deep research Autonomous deep research with structured query elaboration and concise reporting. | 214 | Mar 25, 2025 | MIT | moderate | 40 |
| MCP opennutrition Local search of foods, nutrition facts and barcodes in the OpenNutrition database. | 208 | Jul 28, 2026 | MIT | low | 63 |
| MCP simple arxiv Search and read papers from arXiv. | 201 | Feb 19, 2026 | MIT | low | 48 |
| SerpApi Search results from Google, Bing, Yahoo, DuckDuckGo, YouTube, eBay and more via SerpApi, plus real-time weather and stock market data. | 182 | Oct 15, 2026 | MIT | low | 73 |
| SearxNcrawl Web search via your own SearXNG instance and/or You.com, plus page and site crawling to deduplicated Markdown with Crawl4AI and Playwright, | 181 | Oct 24, 2026 | MIT | moderate | 73 |
| Content core Extract content from web pages, PDFs, Word documents, YouTube transcripts, audio and video with automatic engine selection and structured JS | 181 | Oct 3, 2026 | MIT | low | 68 |
| Agentql MCP ◆ Data extraction with AgentQL. | 181 | Oct 22, 2026 | MIT | low | 78 |
| Keyline Icons Search Keyline Icons by name and get any icon's SVG source or React import. No key, no network. | 178 | Oct 20, 2026 | MIT | low | 74 |
| MCP simple pubmed Search and read medical and life sciences papers from PubMed. | 171 | Mar 19, 2026 | MIT | low | 47 |
| Scrapeless MCP server Web search via the Scrapeless Google SERP API. | 169 | Sep 8, 2026 | MIT | moderate | 67 |
| MCP read website fast Convert websites to clean Markdown with links preserved, using Mozilla Readability, caching, robots.txt-aware crawling and concurrent fetchi | 163 | Sep 29, 2026 | MIT | low | 67 |
| Zotero MCP (kujenga) Search your Zotero library, retrieve item metadata, and read full text from AI assistants. | 162 | Aug 7, 2026 | MIT | moderate | 62 |
| Servo fetch Chromium-free web content extraction in a single binary: fetch, render, crawl and screenshot pages with the Servo browser engine. | 162 | Oct 24, 2026 | Apache-2.0 | low | 67 |
| Arxiv latex MCP Get the LaTeX source of arXiv papers to handle mathematical content and equations. | 148 | Oct 23, 2026 | MIT | low | 67 |
| Web scout MCP Search the web and extract clean, readable text from webpages. Process multiple URLs at once to sp… | 138 | Oct 22, 2026 | Apache-2.0 | low | 72 |
| MCP searxng Connect to SearXNG instances. | 131 | May 29, 2026 | MIT | low | 54 |
| Brave search MCP Web, image, news, video and local point-of-interest search using Brave's Search API. | 130 | May 28, 2026 | GPL-3.0 | low | 54 |
| Google news MCP server Google News via SerpAPI: search headlines, stories and related topics, with automatic topic categorization and multi-language support. | 129 | Oct 20, 2026 | MIT | low | 66 |
| Vectorize MCP server Retrieval, private deep research, file-to-Markdown extraction and text chunking via Vectorize. | 115 | Oct 24, 2026 | MIT | low | 66 |
| Agent search MCP Search across DuckDuckGo, Sogou, Bing, Baidu, Wikipedia, Startpage, Yandex and Mojeek with waterfall search, multi-source verification and n | 112 | Sep 21, 2026 | Apache-2.0 | moderate | 65 |
| MCP screenshot website fast Capture website screenshots, tiling full pages into 1072x1072 chunks for Claude Vision, with configurable viewports and wait strategies for | 110 | Sep 29, 2026 | MIT | low | 65 |
| Doc scraper Crawl documentation sites into local Markdown or JSONL corpora and serve them with offline BM25 search, page reads, freshness checks and cra | 107 | Oct 24, 2026 | Apache-2.0 | low | 70 |
| Watermelon UI Search free React, Tailwind, and shadcn components, blocks, and dashboards; get installable source. | 97 | Oct 24, 2026 | MIT | low | 71 |
| Openai websearch MCP Expose OpenAI's built-in web search tool. | 94 | Sep 12, 2025 | MIT | moderate | 37 |
| Octagon deep research MCP ◆ Deep research agent from Octagon. | 94 | Feb 9, 2026 | MIT | low | 55 |
| Freeq Read and verify freeq conversations: channels, history, search, signatures, pins. | 88 | Oct 24, 2026 | MIT | low | 64 |
| Naver search MCP Naver Search API integration: blog, news and shopping search plus DataLab analytics. | 87 | Oct 7, 2026 | MIT | low | 64 |
| Pdfmux PDF extraction router that classifies each page and routes it to PyMuPDF, Docling, OCR or an LLM fallback, re-extracting pages with low conf | 84 | Oct 19, 2026 | MIT | moderate | 69 |
| MCP zoomeye Query network asset information from ZoomEye. | 82 | Sep 9, 2026 | — | moderate | 49 |
| Plumb MCP Two-way Figma MCP: extract + verify design-to-code, or generate on-brand Figma pages from a prompt. | 81 | Oct 14, 2026 | MIT | moderate | 64 |
| Bing search MCP Web search using the Microsoft Bing Search API. | 81 | Apr 4, 2025 | MIT | moderate | 36 |
| MCP hn Search Hacker News, get top stories and more. | 76 | Jul 14, 2025 | MIT | moderate | 36 |
| Reflex Local, offline full-text code search for coding agents using a trigram index over large codebases. | 76 | Oct 2, 2026 | MIT | low | 64 |
| CN Scraper MCP Search and extract data from Chinese platforms including Taobao, JD, Xiaohongshu, Zhihu, Weibo, Bilibili, Douban and Dianping, with local co | 70 | Jul 29, 2026 | MIT | low | 63 |
| Web Researcher MCP Web search via Google, Brave, Serper, SearXNG and SearchAPI.io, multi-tier content extraction, academic and patent search, and multi-source | 70 | Oct 22, 2026 | MIT | moderate | 68 |
| Linkedapi MCP ◆ Control LinkedIn accounts and retrieve real-time LinkedIn data. | 70 | Oct 16, 2026 | MIT | low | 73 |
| Token Enhancer Cuts AI web-fetching costs up to 99.9% by stripping page junk before it reaches your LLM. | 68 | Apr 5, 2026 | MIT | low | 43 |
| Texas grocery MCP MCP server for HEB grocery store integration - search products, manage carts, clip coupons, and more | 67 | May 11, 2026 | MIT | low | 51 |
| Lilbee Index files, code and crawled websites and answer with source citations, using its own local models or an existing Ollama or LM Studio setup | 66 | Oct 24, 2026 | MIT | low | 63 |
| Mobbin Search real-world UI & UX design references for mobile apps, web apps, and websites with Mobbin. | 65 | Sep 24, 2026 | MIT | low | 64 |
| Job searchoor Search job listings with filters for date, keywords, remote work options and more. | 65 | Apr 8, 2025 | MIT | moderate | 35 |
| MCP domain availability Check domain availability across many TLDs with DNS and WHOIS verification, bulk checking and name suggestions. | 61 | Sep 27, 2026 | MPL-2.0 | low | 63 |
| Neuronto ARD Registry: Agentic Resource Discovery (ARD) Index Neuronto ARD Registry: one search across every public ARD registry + 200,000+ verified MCP tools | 56 | Oct 3, 2026 | Apache-2.0 | low | 68 |
| SiGit MCP server for siGit (sigit.si): browse repos, search code, manage PRs/issues, web search. | 56 | Oct 22, 2026 | — | moderate | 58 |
| Webpage screenshot MCP Take screenshots of webpages to use as feedback during UI development. | 56 | Jun 29, 2025 | MIT | moderate | 34 |
| Sifter Extract typed, schema-defined records from any document and query them in natural language like a database. | 54 | Sep 17, 2026 | MIT | low | 57 |
| MCP server tavily Search via the Tavily AI search API. | 50 | Aug 19, 2025 | MIT | moderate | 34 |
| MCP server webcrawl Search and retrieve web crawler data from WARC, wget, Katana, SiteOne and InterroBot crawls. | 46 | Sep 28, 2026 | — | moderate | 52 |
| WebScraping.AI ◆ Web data extraction and scraping via WebScraping.ai. | 45 | Oct 24, 2026 | — | high | 67 |
| Competitor Tracking Trends MCP Live trend data from Google, TikTok, YouTube, Amazon, Reddit, npm, Steam and other sources. | 44 | Aug 29, 2026 | MIT | low | 67 |
| MCP servers kagi Kagi search API integration. | 44 | Dec 13, 2024 | MIT | moderate | 33 |
| Dappier MCP Real-time web search and premium data from media brands via Dappier: news, financial markets, sports, entertainment, weather and more. | 44 | Jun 27, 2025 | MIT | moderate | 33 |
| Kagi ken MCP Kagi search and summarization without API access, authenticated with a Kagi session token. | 42 | Feb 15, 2026 | MIT | low | 41 |
| Bilibili MCP Server Local-first Bilibili MCP server for video metadata, subtitles and structured transcripts, chapters, popular comments, and optional local ASR | 40 | Oct 23, 2026 | Apache-2.0 | low | 66 |
| Baseline MCP server Search Baseline status using the Web Platform API. | 37 | Oct 21, 2026 | MIT | low | 61 |
| Newsmcp Real-time world news: events clustered from many sources, classified by topic and geography and ranked by importance. | 35 | Mar 6, 2026 | MIT | low | 45 |
| Unsplash MCP Unsplash photo search, random photos and download tracking, with ready-to-use attribution text and HTML for each photo. | 34 | Jan 10, 2026 | MIT | low | 40 |
| Youtube MCP Transcribe YouTube videos to text using yt-dlp for audio and OpenAI Whisper for transcription, with chunked output for long videos. | 32 | Jul 21, 2025 | — | high | 22 |
| Domain search MCP Domain availability and prices from Porkbun, Namecheap, GoDaddy, RDAP and WHOIS, with bulk search, registrar comparison, suggestions and soc | 28 | Oct 5, 2026 | MIT | moderate | 64 |
| Stocky Search and download royalty-free stock photos from Pexels and Unsplash, with rich metadata and pagination. | 28 | Oct 19, 2026 | MIT | low | 59 |
| Parallel Search MCP Web search for AI via Parallel. | 27 | Oct 5, 2026 | MIT | low | 70 |
| Searchpin — Free Web Search for AI Agents Multi-engine parallel web search with re-ranking that works within China's network without a proxy or VPN. | 27 | Oct 23, 2026 | — | moderate | 54 |
| MCP rss aggregator Aggregate RSS feeds. | 27 | Oct 17, 2026 | MPL-2.0 | low | 59 |
| MCP paperswithcode Search papers via the Papers with Code API. | 26 | Jun 7, 2025 | MIT | moderate | 31 |
| Json MCP filter Query and extract only the needed parts of JSON files to keep LLM context small. | 26 | Mar 30, 2026 | MIT | low | 39 |
| Overseerr MCP Media discovery, requests and management via Overseerr and Seerr for Plex, Jellyfin and Emby. | 25 | Oct 21, 2026 | MIT | low | 59 |
| Opentk MCP Search Dutch Parliament (Tweede Kamer) documents, debates, activities and legislative cases, based on the OpenTK project. | 23 | Oct 30, 2025 | MIT | low | 39 |
| Nexus Web search using Perplexity Sonar models, with source citations. | 22 | Oct 21, 2026 | MIT | low | 58 |
| Himalayas Remote Jobs Real-time access to remote job listings and company information from the Himalayas jobs database. | 21 | Oct 21, 2026 | MIT | low | 69 |
| HasData Structured data APIs for Google Search, Maps, Trends and Flights, Amazon, Airbnb, Zillow, Yelp and more, returning clean JSON. | 20 | Oct 12, 2026 | MIT | low | 69 |
| StitchAPI Docs Semantic search over the StitchAPI documentation, returning relevant sections with deep links and full pages. | 20 | Oct 22, 2026 | Apache-2.0 | low | 69 |
| Domain suite MCP Domain lifecycle management across Porkbun, Namecheap, GoDaddy and Cloudflare: availability, registration, DNS, SSL, email auth (SPF, DKIM, | 20 | Mar 12, 2026 | MIT | low | 38 |
| Nyt Search articles using the NYTimes API. | 18 | Jul 23, 2026 | MIT | low | 46 |
| GeekNews MCP Server Retrieve and process news from GeekNews. | 18 | Apr 13, 2025 | MIT | moderate | 30 |
| MCP Retrieval Read-only web search via DuckDuckGo Lite, image search via Bing Images and page-to-Markdown scraping with a readability extractor. | 17 | Oct 24, 2026 | MIT | moderate | 62 |
| Activitypub MCP Explore and interact with the Fediverse via ActivityPub: WebFinger discovery, timelines and instance exploration across Mastodon, Pleroma, M | 17 | Sep 13, 2026 | MIT | low | 57 |
| Parallel Task MCP Deep research and batch task execution via Parallel. | 16 | Sep 11, 2026 | — | moderate | 53 |
| Nab Fetch any URL as clean Markdown, with HTTP/3, JS rendering, anti-fingerprinting, browser cookie auth and 1Password integration. | 16 | Oct 24, 2026 | MIT | low | 62 |
| Gopher & Gemini MCP Server Browse Gopher and Gemini protocol resources, with TLS support and structured content extraction. | 16 | Oct 19, 2026 | MIT | low | 62 |
| Webcrawl MCP Local-first web scraping, DuckDuckGo search, site mapping and crawling, using trafilatura with an optional Firecrawl fallback for JS-rendere | 16 | Sep 9, 2026 | MIT | low | 52 |
| Multi research agents Multi-agent research system for coding or general research tasks. | 16 | Mar 19, 2026 | MIT | low | 37 |
| Enrichr MCP server Gene set enrichment analysis using the Enrichr API. | 15 | Jul 14, 2026 | — | moderate | 35 |
| Ncbi MCP server NCBI and PubMed literature search with analytics, caching, MeSH integration, related article discovery and batch processing for life science | 13 | Jun 28, 2025 | — | high | 18 |
| Dreamwork ◆ Dreamwork job search: browse live listings without a key, rank matches against your resume, generate tailored resumes and cover letters, app | 12 | Oct 23, 2026 | MIT | moderate | 71 |
| SearXNG HTTP MCP Self-contained SearXNG in Docker: metasearch across many engines and categories with multi-page fanout, autocomplete, engine discovery and a | 12 | Oct 23, 2026 | MIT | moderate | 61 |
| Melrose MCP Play Melrōse music expressions as MIDI. | 12 | Sep 21, 2026 | MIT | low | 51 |
| MCP tavily Search via the Tavily AI search API. | 12 | Jun 9, 2026 | MIT | low | 44 |
| Driflyte MCP server ◆ Query and retrieve topic-specific knowledge from recursively crawled and indexed web pages via Driflyte. | 11 | Oct 2, 2025 | MIT | moderate | 42 |
| HasData Indeed Remote MCP server for Indeed: job listings by keyword and location with salary, company and posting date, plus full job postings, as JSON. | 11 | Oct 12, 2026 | MIT | moderate | 66 |
| Crawleo MCP Real-time web search and crawling via the Crawleo Search and Crawl API. | 11 | Dec 31, 2025 | MIT | low | 40 |
| Agent Web Search Provider-neutral web search aggregating model grounding (Gemini, Grok, DeepSeek and others) and search APIs such as Exa, Brave, Perplexity a | 11 | Oct 23, 2026 | MIT | moderate | 60 |
| HasData Web Scraping Web scraping: fetch public pages with optional JavaScript rendering, wait conditions and CSS or AI extraction rules, as Markdown, text, HTML | 10 | Oct 12, 2026 | MIT | moderate | 66 |
| Decompose Deterministically decompose text into classified semantic units with authority, risk and attention scores plus entity extraction, without an | 10 | May 24, 2026 | MIT | low | 48 |
| MinerU Parse PDFs, images, DOCX and PPTX via the MinerU document parsing API, with multilingual OCR, batch processing, page ranges and local file u | 10 | Oct 19, 2026 | MIT | low | 60 |
| Heventure search MCP Web search across DuckDuckGo, Bing, Google (via SerpAPI) and Tavily with automatic rate limiting, caching and multi-language support. | 10 | May 13, 2026 | — | moderate | 33 |
| MCP claude hackernews Interact with Hacker News. | 10 | Oct 21, 2026 | MPL-2.0 | low | 55 |
| HasData Bing Bing search results as JSON: organic results, ads and the Copilot answer, targeted by market, country, language and device. | 9 | Oct 12, 2026 | MIT | moderate | 66 |
| Citation Intelligence Check which URLs Perplexity, Claude, ChatGPT, Gemini, Bing and Google AI Overviews cite for any query. Self-hosted. | 9 | Oct 13, 2026 | MIT | low | 60 |
| MCP pearch ◆ People search engine for talent discovery. | 9 | Oct 20, 2026 | MIT | low | 65 |
| Searchcraft MCP server ◆ Manage Searchcraft clusters, create search indexes, generate an index from a data file and import data from a feed or local JSON file. | 9 | Feb 7, 2026 | Apache-2.0 | low | 45 |
| Shopsavvy MCP server ◆ ShopSavvy product and price data: search by barcode, ASIN or URL, get product details, compare retailer prices and track price history. | 9 | Oct 12, 2026 | MIT | low | 65 |
| HasData Google Images Remote MCP server for Google Images: image results with source pages, thumbnails and full-resolution URLs, filtered by size, colour and type | 8 | Oct 12, 2026 | MIT | moderate | 65 |
| Gsc MCP Google Search Console and Indexing API: search analytics, sitemaps, URL inspection and batch indexing. | 8 | Oct 19, 2026 | MIT | moderate | 59 |
| Scavio Unified real-time search across Google, YouTube, Amazon, Walmart, Reddit and TikTok for web, e-commerce, product, social media and video dat | 8 | Oct 23, 2026 | MIT | moderate | 59 |
| Boikot Look up company ethics information: the ethical and unethical actions of major companies. | 8 | Oct 24, 2026 | GPL-3.0 | low | 54 |
| Fouradata MCP ◆ Web scraping that escalates from direct fetch to rotating proxy to full browser, solving anti-bot challenges and reporting which step worked | 7 | Oct 24, 2026 | MIT | moderate | 68 |
| Argus Retrieval Multi-provider search broker with automatic fallback, RRF ranking, content extraction and budget enforcement. | 7 | Oct 22, 2026 | MIT | moderate | 58 |
| Google AI Search MCP Search, documentation retrieval, code analysis and architecture research powered by Google's Vertex AI or the Gemini API. | 7 | Sep 7, 2026 | MIT | moderate | 53 |
| Keenable MCP ◆ Live web search and clean Markdown page fetching over the Keenable web index. | 6 | Oct 13, 2026 | MIT | moderate | 68 |
| Youtube summarize Fetch YouTube transcripts in text, JSON, SRT or WebVTT format and in multiple languages, and optionally summarize them. | 6 | Oct 2, 2026 | MIT | low | 58 |
| Clojars MCP Server Up-to-date dependency information for Clojure libraries from Clojars. | 6 | Jun 25, 2025 | MIT | moderate | 25 |
| Agent toolbox Web search, content extraction, screenshots, weather, finance, email validation, translation, news, GeoIP, WHOIS, DNS, PDF extraction and QR | 6 | Mar 10, 2026 | MIT | low | 33 |
| Codearia sieve Deterministically convert web pages into decision-ready state: ISO dates, numbers with units as facts and token-budgeted chunks anchored to | 5 | Sep 22, 2026 | MIT | low | 52 |
| Stock Images MCP Search free stock photos on Pexels, Unsplash and Pixabay and download the chosen one into a local folder, with per-provider error reporting. | 5 | Oct 19, 2026 | MIT | moderate | 57 |
| Anybrowse Scrape, crawl and search the web, converting any URL to LLM-ready Markdown via real Chrome browsers. | 5 | May 26, 2026 | MIT | low | 45 |
| Marketplace search MCP Search TCGPlayer, Reverb and Thumbtack, verify US contractor and nurse licenses, and look up PSA card grading population data. | 5 | Mar 29, 2026 | — | moderate | 27 |
| Talonic Schema-validated extraction of structured fields from PDFs, scans, images and forms, with searchable memory of past extractions. | 5 | Oct 21, 2026 | MIT | moderate | 57 |
| Markcrawl Crawl websites into clean Markdown, search pages and extract structured data with LLMs for web research and RAG pipelines. | 5 | Sep 23, 2026 | MIT | low | 47 |
| NyxDocs Manage cryptocurrency project documentation across Ethereum, BSC, Polygon and Solana. | 5 | Jun 24, 2025 | MIT | moderate | 24 |
| Scraperapi MCP Web scraping via ScraperAPI with JavaScript rendering, geotargeting, premium proxies and auto-parsing. | 5 | Jul 17, 2026 | MIT | low | 40 |
| Web search plus MCP Multi-provider web search with automatic routing between Serper, Tavily and Exa. | 5 | Oct 18, 2026 | MIT | low | 52 |
| String ai MCP ◆ Search the web, fetch URLs and map sites as clean Markdown, getting past anti-bot blocks. | 4 | Oct 18, 2026 | MIT | moderate | 66 |
| Rephonic ◆ Search Rephonic podcast and episode data: listener estimates, demographics, contacts, transcripts, chart rankings, sponsors, reviews and aud | 2 | Oct 23, 2026 | MIT | low | 63 |
| UnblockingAPI ◆ Fetch bot-protected and JavaScript-heavy pages, choosing plain HTTP or a real browser per site, and parse supported pages into JSON with pub | 2 | Oct 22, 2026 | MIT | moderate | 63 |
| Forge MCP ◆ Generate vector embeddings for semantic search and RAG with Forge, Voxell's text-embedding API, including Matryoshka truncation. | 1 | Oct 5, 2026 | MIT | moderate | 60 |
| YouTube Transcript & Search MCP Server ◆ YouTube transcripts (text, JSON, SRT, VTT, srv3), video and channel search, playlists and large batch jobs. | 1 | Oct 15, 2026 | MIT | low | 60 |
| Desearch ◆ AI search, X search and web search for AI agents. | 1 | Oct 22, 2026 | MIT | moderate | 60 |
| Scribefy ◆ Extract timestamped YouTube transcripts, plus video search, metadata and related videos. | 1 | Jun 10, 2026 | MIT | moderate | 48 |
| Muumuu Domain MCP Server ◆ Search and register Muumuu Domain (GMO Pepabo) domains, manage owned domains and contracts, and configure DNS records. | 1 | Aug 4, 2026 | MIT | low | 55 |
| Pod MCP ◆ Search firsthand observations agents recorded about products, APIs, services or places, as opposed to documentation claims, and write back n | 1 | Oct 6, 2026 | MIT | low | 55 |
| Litescrape ◆ Google Search, Bing, DuckDuckGo, Google Maps, Google AI Mode, AI Overview and Shopping results as JSON. | 0 | Oct 19, 2026 | MIT | moderate | 60 |
| SERPdive Web Search ◆ Web search returning extracted page content (URL, title, date, text) instead of links, with automatic localization and an optional synthesiz | 0 | Jul 25, 2026 | MIT | moderate | 48 |
| ContextFetch ◆ Web search and page contents as LLM-ready Markdown, pruned to the passages relevant to the query, plus cited answers, research, crawl and si | 0 | Oct 22, 2026 | MIT | moderate | 60 |
| Parserbee MCP ◆ Extract structured data from invoices, receipts, purchase orders, bank statements and other documents with ParserBee Parser Templates, retur | 0 | Oct 18, 2026 | MIT | moderate | 60 |
| Techmap Job Postings ◆ Search and count job postings from 185 sources in 250 countries (~8M new per month) via Techmap's Jobs API; filters for title, skills, compa | 0 | Oct 1, 2026 | MIT | moderate | 60 |
| Free proxy MCP ◆ Verified free HTTP, SOCKS4 and SOCKS5 proxy lists from Litport, filtered by country, anonymity, latency and uptime. Returns records only; no | 0 | Sep 21, 2026 | MIT | low | 55 |
| Quanticdata MCP server ◆ Web data via residential proxies: page scraping to Markdown, SERP search, site crawl and map, ready-made collectors and datasets from plain- | 0 | Oct 3, 2026 | MIT | moderate | 60 |
| ToolTrace io MCP ◆ Web page data: scrape to Markdown, extract metadata, links and JSON-LD, audit on-page SEO, detect tech stacks and validate XML sitemaps. | 0 | Sep 8, 2026 | MIT | moderate | 55 |
| Quantumproxies MCP ◆ Residential-proxy web access: scrape pages to Markdown, SERP search, site mapping and crawling, ready-made collectors, and proxy endpoints f | 0 | Oct 1, 2026 | MIT | low | 55 |
| Crawdar MCP ◆ Evidence-backed business discovery and lead-list research: search by brief or criteria, inspect qualification evidence and exclusions, and e | 0 | Sep 7, 2026 | MIT | low | 50 |
Official servers
Published by the service's own vendor (per the awesome-mcp-servers list): they usually track the API closely and are better maintained.
- Exa MCP server 5.2K ★
- Trieve 2.7K ★
- MinerU Open MCP 229 ★
- Agentql MCP 181 ★
- Octagon deep research MCP 94 ★
- Linkedapi MCP 70 ★
- WebScraping.AI 45 ★
- Dreamwork 12 ★
- Driflyte MCP server 11 ★
- MCP pearch 9 ★
- Searchcraft MCP server 9 ★
- Shopsavvy MCP server 9 ★
- Fouradata MCP 7 ★
- Keenable MCP 6 ★
- String ai MCP 4 ★
Choosing and securing
Compare search providers, language coverage and the ability to read the pages you need. Check extraction, crawling and source preservation functions. Review maintenance, licensing and the terms of external APIs. When comparing local and remote operation, consider cache storage and query confidentiality. Also inspect access limits and error handling so that missing content is visible rather than silently treated as complete.
Retrieved pages can contain malicious instructions, misleading links or unsafe downloads. The server and agent should treat their contents as data without granting them authority. Avoid including secrets in search queries and control access to internal network addresses. Respect source access restrictions. Check links, dates and retrieved passages before sharing a synthesis or relying on its claims for further work.
Before connecting a server to an agent with access to your data, run the AI agent security check.
Most recent
- ContextFetch created October 19, 2026
- Parserbee MCP created October 12, 2026
- Watermelon UI created October 9, 2026
- AffectUI created October 3, 2026
- Dreamwork created October 3, 2026
Other categories
Learn more
Frequently asked questions
What is the best Web search and extraction MCP server?
On our grid (adoption, maintenance, framework), Exa MCP server (97/100), Brave search MCP server (88/100), and Arxiv MCP server (85/100) come first. The right choice mostly depends on the service you need and the rights you are willing to grant.
Are these servers free?
91% declare an open-source licence and are free to use. 43 need an API key: the connected service may be paid.
Can the server read every page on a website?
No. Access depends on the website, technical restrictions and the server's functions. Some pages require authentication or particular rendering capabilities. Check the content actually retrieved and the applicable access conditions. A search result containing a link does not guarantee that the full page text is available.
How can I verify an answer based on web sources?
Request source links and the passages supporting important claims. Open the original pages to check context and dates. Compare evidence when sources disagree. Extraction can omit a table, qualification or separately loaded element, so the retrieved text may not capture everything needed to interpret the page correctly.
Related articles
- Build your own MCP server in 30 minutesOctober 24, 2026
- Automate your inbox with an AI agentOctober 18, 2026
- MCP, A2A, and tool calling: agent protocols explainedOctober 17, 2026
- Build an AI agent with n8n, without codingOctober 16, 2026
Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score