🔧

fetch

Official🔧MCP ServerScanned

modelcontextprotocol/servers

What it does

Web content fetching server that retrieves and processes content from web pages, converting HTML to markdown. Supports chunked reading with start_index for large pages.

Overview

An official MCP server that retrieves and processes web content for LLMs. It fetches web pages, converts HTML to clean markdown, and supports chunked reading for large pages via a start_index parameter.

Key Features

Fetches any web URL and converts HTML content to readable markdown
Handles large pages with chunked reading (start_index pagination)
Strips scripts, styles, and irrelevant elements automatically
Returns clean, structured text suitable for LLM consumption

Who is this for?

Developers who need their AI assistant to read and analyze web content — documentation, articles, API references, or any public web page — directly within the conversation.

webfetchhtmlmarkdownscraping

Installation

Claude CodeAuto-generated from README. View full documentation on GitHub →

claude mcp add fetch -- uvx mcp-server-fetch

Claude Desktop ConfigurationServer configuration (JSON)

{
  "mcpServers": {
    "fetch": {
      "command": "uvx",
      "args": ["mcp-s...

77,494

Last UpdatedJan 27, 2026

View on GitHub Back to MCP Servers