MCP Radar

Fetch MCP Server

The Fetch MCP Server lets your AI fetch a URL and convert the page into clean, model-readable text — an official MCP reference implementation: lightweight and reliable.

For example, "read this doc and summarize the key points" — the AI uses Fetch to grab the page, extract the main text, and summarize.

See this server's health data →

What it does for you

Lets your AI fetch any web page and convert it into readable content — the lightest way to let AI read the web.

Capabilities

  • Fetch URL content
  • Convert to Markdown / plain text

Once installed, try asking your AI

Say this directly in Claude / Cursor or any client with this server connected

  • Read this article and summarize three key points
  • Extract the main text of this page as plain text

Install / Connect

claude mcp add fetch -- npx -y @modelcontextprotocol/server-fetch

Follow the official README; env and args vary per server.

Why use it

Good for: giving the AI a simple, reliable "read a web page" capability without complex crawling.

Maintained by MCP officially; full health data is shown below.

Similar / alternative servers

Frequently Asked Questions

How is it different from Firecrawl?

Fetch is lighter — grabs a single page as text; Firecrawl does more (full-site crawl, search). For simple page reads, Fetch is enough.

Can it fetch pages behind a login?

It mainly fetches public pages; for complex auth, use a more specialized scraping server.