fetch

Overview

The fetch MCP server provides a simple yet powerful tool for LLMs to retrieve and process web content. It converts HTML to markdown for easier consumption by language models, with options to control content length and starting position. The server includes security considerations, warning users about potential access to local/internal IP addresses. It supports various installation methods including uv, pip, and Docker, making it accessible across different environments.

Try asking AI

After installing, here are 6 things you can ask your AI assistant:

• fetchurl: string, max_length?: integer, start_index?: integer, raw?: boolean

  Fetches a URL from the internet and extracts its contents as markdown

  Try:Fetch https://example.com and convert to markdownTry:Get the content from https://news.example.com starting from character index 1000Try:Fetch https://example.com with raw HTML output

Comparable tools

Installation

uvx mcp-server-fetch

pip install mcp-server-fetch

python -m mcp_server_fetch

{
  "mcpServers": {
    "fetch": {
      "command": "uvx",
      "args": ["mcp-server-fetch"]
    }
  }
}

FAQ

Can the fetch server access local/internal IP addresses?

Yes, the server can access local/internal IP addresses which may represent a security risk. Use caution to avoid exposing sensitive data.

How can I customize the user-agent for fetch requests?

Add --user-agent=YourUserAgent to the args list in your configuration to customize the user-agent string.

On Hacker News

Recent discussion from the developer community.

• Modelcontextprotocol/Servers: Model Context Protocol Servers▲ 1
  Story by vinnyglennon · 2025-03-20

Compare fetch with