Smithery is now a part of Arcade.dev! Read more in our announcement here
Search the web with Brave's independent index — web, news, images, and videos. Bring your own subscription token from the Brave Search API dashboard.
AI Search (AutoRAG) management. List and search documents across your AI Search indexes.
Fast, intelligent web search and web crawling. Get fresh information about libraries, APIs, and SDKs.
Find and download academic papers from leading sources like arXiv, PubMed, bioRxiv, medRxiv, Google Scholar, Semantic Scholar, CrossRef, and IACR. Get standardized results and fetch full-text PDFs when available. Accelerate literature reviews with deep search and effortless retrieval.
Docs: https://docs.keenable.ai/mcp-server Keenable is a free, remote MCP server that gives agents access to the web index. Search the web with ranked results and date/site filters, then fetch any indexed page as clean markdown. Works out of the box with no account or API key. Tools: searchwebpages: Search the web and return ranked results. fetchpagecontent: Fetch an indexed URL and extract its content as clean markdown
Search the Daum index across web, video, image, blog, book, and cafe verticals. Pairs with Naver Search for broader Korean-web coverage; Daum's video and cafe surfaces are especially distinct.
GroundRoute gives your AI agent web search across six engines (Serper, Brave, Exa, Tavily, Firecrawl, Perplexity) through a single search tool. That one tool covers web, news, and academic search, synthesized answers with citations, and full-page fetch, all selectable with a mode parameter, so you never wire up or pay for six separate APIs. It routes each query to the cheapest engine that meets a quality bar, caches repeats, and fails over automatically. Pricing is gain-share: you keep about half the cache savings, so it never costs more than going direct. Bring your own engine keys (BYOK) or let GroundRoute manage them. New accounts start with $10 in free search credit, no card. Get a key at https://groundroute.ai/keys, and see the live benchmark across all six engines (170 real agent queries) at groundroute.ai/state-of-ai-search.
Highest accuracy web search for AIs
A free, open MCP server for The Urantia Papers — 197 papers, 1,626 sections, 14,500+ paragraphs, 4,400+ entities. No authentication required. Tools (13) search / semanticsearch — full-text + vector similarity search across all paragraphs getparagraph / getparagraphcontext — look up by reference (auto-detects 3 ID formats), with optional surrounding context getpaper / getpapersections / gettableofcontents — paper-level navigation listentities / getentity / getentityparagraphs — browse 4,400+ entities (beings, places, concepts, etc.) getrandomparagraph — daily quotes, exploration getaudio — paragraph-level TTS audio URLs Resources urantia://paper/{id} — full paper rendered as markdown for RAG/summarization urantia://entity/{id} — entity description with all paragraph references Prompts studyassistant — primes the model as a Urantia Papers study guide (optional topic arg) comparative_theology — struct
Advanced patent search across 200M+ records using semantic, similarity, image, patent-number, nested-query, analytics, and keyword-assist workflows.
Web-scale search for AI thats 100x cheaper and 10x faster. Search More. Pay Less.
Search Yelp for local businesses and reviews. No API key required.
Search Tripadvisor for hotels, restaurants, and attractions. No API key required.
Search Naver across news, blogs, books, encyclopedia, cafe posts, Knowledge iN, local places, images, shopping, and professional documents. Returns Korean-language results and Korea-local content that global search engines often miss.
Search Switzerland's official building applications for new residential projects, months before they are advertised. Zuerst zuhause reads every building application published in Switzerland (cantonal gazettes, municipal notices, cantonal building portals), every day, and turns them into structured project records: what is being built, where, how many apartments, who the builder is, when the objection period ends, with a link to the original publication. Tools searchprojects: filter by municipality, canton, free text, category (new build, conversion, renovation…), market type (buy/rent), minimum units, publication window; paged, up to 50 per call. getproject: one project by id. Access https://zuerstzuhause.ch/mcp: no authentication; project description and place, fresh projects as teasers. https://zuerstzuhause.ch/mcp/konto: OAuth 2.1 with dynamic client registration and PKCE; subscribers get address, builder, architect, zone and source link, plus saved-sear
Alexandria gives AI agents semantic search over 4.6 million text chunks from 20,000+ classical philosophy and humanities works — Plato, Aristotle, Kant, Nietzsche, Hegel, Hume, Descartes, Schopenhauer, and thousands more — sourced from Archive.org. Texts span Ancient Greek, Latin, German, French, English, Italian and other languages. Use searchtexts to find passages by concept, argument or author. Use getbooklist to browse available works. Three prompts are available for guided analysis: philosopheranalysis, topicexploration, and comparephilosophers.
BOE, Spanish public tenders and INE statistics behind one MCP server. Four tools over the official APIs: boesearch, boeday, tenderssearch, ineseries. Every answer carries its source URL. A free key answers 100 calls a day; get one at https://automate-it-all.win/datos/.
Unified real-time search engine skill for AI agents.
MCP server with EU public procurement data: search open tenders, see who won similar contracts and at what price, and get bid intelligence before you bid.
Full-text retrieval for AI agents over peer-reviewed papers, books, patents, Wikipedia, Reddit, Telegram, and YouTube. Four read-only tools: searchdocuments searchsocial fetchdocument searchin_document — return canonical source URIs (DOI / arXiv / PMID / ISBN) for verbatim citation.
