Agent-optimized MCP server for extracting clean, structured content from web pages - built specifically for AI agents that need
294 servers · 294 listings
Uses Playwright to crawl web content for AI models.
MCP Server for searching via DuckDuckGo (Node.js version)
MCP server for the EnConvert web data and file conversion API: perceive pages (markdown, HTML, screenshots, PDF), discover URLs, search the
MCP server for NerdyData API — search websites by source code, technology stack, analyze domains, and download lead lists through LLM chat
Fetches web content with automatic image saving and optional AI display.
Apify MCP Server
MCP server that runs Google AI Mode searches via a browser sidecar extension using your own logged-in session.
A Model Context Protocol (MCP) server that provides AI-powered Google search using Gemini with search grounding
Unified search across DuckDuckGo, GitHub, npm, PyPI, Stack Overflow, Wikipedia and Hacker News — for AI agents. No API keys required. Part
Give your AI agent the ability to extract clean, structured data from any web page. Articles, products, metadata, tables, contact info
MCP server giving any AI agent pay-per-call web search, settled in USDC on Base via x402. Wraps the Base Intel /search API. The agent pays
MCP 服务器,提供中文必应搜索结果抓取和网页内容抓取功能,支持中文搜索
DeepSearch MCP Server with Brave Search API and Puppeteer content extraction
Research infrastructure for AI agents. Renamed to browseai-dev — installing browse-ai automatically installs browseai-dev. MCP server, REST
A local MCP stdio server that provides access to the OpenAI (ChatGPT) API with web search capabilities for Claude Code and other MCP
MCP fetch server for making HTTP requests
A comprehensive package for searching the web and fetching web content programmatically, through a CLI interface, or through a Model
Dead link checker MCP server - find broken links, redirects, and timeouts on any website. Free tier: 30 links/scan.
MCP server providing web search via DeepSeek's native web_search API
DuckDuckGo search implementation for Model Context Protocol (MCP)
Web, Search & Scraping MCP servers on mcprush read the open web on the agent’s behalf: pages fetched and turned into text it can cite, feeds watched for the changes you named, locations and routes answered in one call rather than ten.
What to look at before installing one is what the allowance is and how it is counted. A metered server prints its own ceiling, and the gateway counts against it where you can see the number. Every server here names the publisher behind it and whether that publisher has been verified, the runtime it uses, how it authenticates, the tool surface it exposes and the price model it is sold under — and where a security scan has been run against it, what the scan found. The filters on the left narrow it further and the sort at the top decides what comes first.
A server that runs on its publisher’s infrastructure is reached through the mcprush gateway, so one key covers everything installed against your account, the calls are metered where you can see them and the ceiling on what they may cost is yours to set. Installing one is a line of the mcprush CLI — npx mcprush@latest add, followed by the server’s id — and a --client flag writes another client’s config instead of the default one. A server that runs on your own machine carries its publisher’s own instructions instead, and the listing says which of the two it is before you install anything.