RAG Web Browser
FreeApify 开源工具，执行网页搜索、抓取 URL 并以 Markdown 格式返回内容。
FreeFree tier
Outputs: text
About RAG Web Browser
An MCP (Model Context Protocol) server that enables LLMs and AI agents to perform web searches and fetch URLs. It communicates with Apify's RAG Web Browser Actor in Standby mode, providing fast responses. Supports Google search queries with configurable number of results, direct URL fetching, and returns cleaned content as Markdown, text, or HTML. Offers two scraping modes: raw-http for fast static pages and browser-playwright for JavaScript-heavy sites. Note: This server is deprecated in favor of mcp.apify.com, which provides the same functionality plus access to 6,000+ Apify Actors.
Key Features
Web search via Google
Single URL fetching
Returns content as Markdown, text, or HTML
Two scraping modes: raw-http and browser-playwright
Configurable max results, timeout, and scraping tool
Designed for fast LLM responses
Pros & Cons
Pros
- Open source and free to use
- Supports both search and direct URL fetching
- Returns clean, structured content in multiple formats
- Configurable scraping mode (fast vs JS-heavy)
- Works with any MCP-compatible client (Claude, VS Code, etc.)
Cons
- Deprecated in favor of centralized mcp.apify.com
- Requires local setup and Node.js/npx
- Limited to 1 result by default, may need tuning
- Only supports Google search
Best For
Providing LLMs with real-time web search resultsFetching and cleaning web page content for RAG pipelinesIntegrating web browsing into AI agents via MCP
FAQ
What does this MCP server do?
It allows AI agents and LLMs to search the web or fetch specific URLs, returning cleaned content as Markdown, text, or HTML.
Is this server deprecated?
Yes, it is deprecated in favor of mcp.apify.com, which provides the same functionality plus access to 6,000+ Apify Actors.
How do I use it?
You can run it locally using npx with the command `npx @apify/mcp-server-rag-web-browser`, or migrate to mcp.apify.com for easier setup.