Hosted tools for agents

MCP servers, one URL away.

A hosted MCP server gives your agent live tools and data without a local install. Browse 3,808 MCP servers, copy the endpoint, and use the one-line installer for your client.

MCP servers
3,808
Categories
15
Client guides
14
Distinct hosts
3,667

Showing 0–24 of 109 servers

Clear filters
AhrefsOfficialWeb ScrapingOfficial Ahrefs MCP server that lets AI agents securely access Ahrefs SEO data, site metrics, keyword research, and competitor analysis through the Ahrefs API.api.ahrefs.comView setup2CaptchaOfficialWeb ScrapingOfficial 2Captcha MCP server: web search, scraping past anti-bot protection, marketplace product/search parsing into structured JSON, browser automation and CAPTCHA solving for AI agents. Local (npx @2captcha/mcp) or remote (https://mcp.2captcha.com/mcp).mcp.2captcha.comView setupApifyOfficialWeb ScrapingConnects AI agents to thousands of web scraping, data extraction, and automation tools available on the Apify Store.mcp.apify.comView setupBrowser UseOfficialWeb ScrapingOfficial Browser Use MCP server: run browser agents on stealth cloud browsers (CAPTCHA solving, proxies in 195+ countries) from Claude Code, Cursor or any MCP client.api.browser-use.comView setupCaptapiOfficialWeb ScrapingPublic social media data for AI agents: profiles, posts, comments, transcripts and search from TikTok, Instagram, YouTube, X, Facebook, Threads, Reddit and LinkedIn company pages, plus Meta, Google, LinkedIn and TikTok ad libraries.api.captapi.comView setupFastSocial Instagram DataOfficialWeb ScrapingRead-only public Instagram data as JSON: profiles, posts, reels, stories, comments and daily top-accounts charts. Free plan with 50 credits a month.data.fastsocial.coView setupFirecrawlOfficialWeb ScrapingIntegrates web scraping and content extraction into LLM clients like Cursor and Claude. Three connection paths: keyless at https://mcp.firecrawl.dev/v2/mcp for search, scrape and parse within daily limits; browser OAuth at https://mcp.firecrawl.dev/v2/mcp-oauth; or an API key on the standard endpoint.mcp.firecrawl.devView setupLinkedIn DataOfficialWeb ScrapingLook up people, companies, jobs and posts, and find and verify work emails.api.envoapi.comView setupReduckOfficialWeb ScrapingReduck MCP allows agents to discover, run and create browser automation scripts that serve as tools. Scripts run in your own Chrome, through our extension, which allows your agent to work where you are logged in — no credentials exposure, and same fingerprints so no bot detection.mcp.reduck.aiView setupZyteOfficialWeb ScrapingOfficial Zyte MCP server to give you agent web access via Zyte API, extract data, render pages, and manage your Scrapy Cloud projects.mcp.zyte.comView setup2extractWeb ScrapingCreate and manage residential and mobile proxies from your AI agent: geo targeting down to city, ZIP or carrier, traffic and spend limits, usage and cost reportsmcp.2extract.comView setupAdam Hale Tools: Web Research and Open DataWeb ScrapingRemote MCP server (Streamable HTTP, no account): website tech stack, sitemap URLs, RSS to JSON and EU TED tenders free for single items; page to Markdown, domain WHOIS/RDAP + DNS, robots.txt checks and bulk calls paid per call in USDC via x402.tools.adamhaleworks.workers.devView setupagdataWeb ScrapingPay-per-call web data for AI agents: any URL as Markdown, screenshots, YouTube transcripts, Google Maps and Amazon reviews, job listings, news and search. No API key, paid per call with x402 stablecoins.agdata.shveik.devView setupAgent Web ReaderWeb ScrapingWeb search or URLs in, clean markdown out, one JSON row per page with source URL, title and status. $0.002 per page; a failed URL is never charged and a per-URL error is returned as a row instead of failing the call. Token budget per result, HTML and PDF, built for agents. Seven tools, hosted on Apify.mcp.apify.comView setupAgentDiet - Token Diet Semantic De-bloaterWeb ScrapingStrips 85-97% of HTML token bloat for AI agents with 4 modes (skeleton recon, BM25 token budget, Markdown, JSON tables), using Bearer API keys or HTTP 402 (x402/L402) micropayments.agentdiet-72343955726.us-central1.run.appView setupAgentKit x402Web ScrapingPay-per-call utility API for agents over x402 (USDC on Base): email and domain checks, DMARC audit, site audit, URL to markdown, and IBAN validators. Pricing and demo routes are free; other tool calls return HTTP 402 until the client signs an x402 v2 payment.agentkit-x402.zenith-ops.workers.devView setupAgentReadWeb ScrapingWeb page to clean Markdown, metadata, headings and links for AI agents. Free inspect; paid extraction via x402 (USDC on Base).agentread.omar-cv.workers.devView setupAgentReaderWeb ScrapingTurns any URL or PDF into clean Markdown or JSON for AI agents. Remote MCP exposes read_url, search, get_balance, and report_issue, plus a REST API. Five free reads a day per IP with no signup, then credits or $0.005 per call via x402 USDC on Base.agentreader.xyzView setupAgentRenderWeb ScrapingURL to screenshot, PDF or structured extract for AI agents. Tools: screenshot, pdf, extract and get_free_key. ChatGPT can sign in through OAuth at https://api.agentrenderapi.com/mcp with no key to paste; Claude, Cursor and other clients run uvx agentrender-mcp with a free 150-render key, obtainable from the provider or through get_free_key.api.agentrenderapi.comView setupagentsvc.ioWeb Scraping127 pay-per-call services for AI agents: web pages to Markdown, screenshots, PDF, OCR, web and discussion search, company and compliance checks, crypto and market data, DeFi yields, validation and charts. Paid per call in USDC on Base, Solana, Polygon or Arbitrum via x402 inside the MCP call. No account, no API key, and every service has a free trial endpoint.agentsvc.ioView setupagproxyWeb ScrapingPay-per-call residential proxy traffic and raw fetch for AI agents, with country targeting and sticky sessions. No API key, paid per call with x402 stablecoins.agproxy.shveik.devView setupAllscreenshotsWeb ScrapingScreenshot API for full-page and viewport captures as PNG, JPEG, WebP or PDF, plus HTML, Markdown and JSON extraction from a single page load. Supports device emulation, dark mode, ad and cookie-banner blocking, scheduled captures, bulk capture, official SDKs and a hosted MCP server. Remote MCP endpoint (OAuth): https://mcp.allscreenshots.com/mcp/v1. Docs: https://docs.allscreenshots.com. 100 free lifetime credits; paid plans from $10/month.mcp.allscreenshots.comView setupAnakinWeb ScrapingAnakin is a web data infrastructure layer for AI: search, scrape, crawl and monitor the live web without maintaining browsers, proxies or scraping infrastructure.mcp.anakin.ioView setupBowmarkWeb ScrapingBowmark turns websites into typed functions an AI agent can call. Read a library, write a short script, get structured data back. Every request needs a Bowmark API key.api.bowmark.aiView setup

Endpoints and setup links are checked against provider documentation where available. Verify access, scopes, and rate limits with the provider before connecting.