fetch-mcpA flexible HTTP fetching Model Context Protocol server.

— (0)

Reviewed by Daniel Nikulshyn·Updated July 2026

Overview

Fetch MCP Server is an MCP server for fetching web content in multiple formats. It allows users to fetch web content as HTML, JSON, plain text, Markdown, readable article content, and YouTube transcripts. The server accepts common parameters such as URL, custom headers, maximum length, start index, and proxy URL. It also supports various commands for different types of fetches, including fetch_html, fetch_markdown, fetch_txt, fetch_json, fetch_readable, and fetch_youtube_transcript. The tool is designed to be flexible and can be used as an MCP server or as a CLI tool.

Key features

Fetch web content as HTML, JSON, plain text, or Markdown
Extract article content with Mozilla Readability
Extract YouTube video transcripts
Proxy support for requests
Pagination with max_length and start_index
Custom request headers

Pricing

Model: Free
Category: MCP Servers
Rating: No reviews yet

Use cases

Fetching Article Content

Use the fetch_readable command to extract the main article content from a webpage, stripping away navigation, ads, and boilerplate.

Extracting YouTube Transcripts

Use the fetch_youtube_transcript command to extract the transcript of a YouTube video, with optional language selection.

Fetching Web Content as Markdown

Use the fetch_markdown command to fetch a webpage and return its content converted to Markdown.

Pros & Cons

Pros

Flexible and supports multiple formats for fetching web content
Supports extraction of article content using Mozilla Readability
Can extract YouTube video transcripts
Has features like proxy support, pagination, and custom request headers
Includes security features like SSRF protection and response size limits

Cons

Requires knowledge of MCP client configuration for server setup
Some features like YouTube transcript extraction require additional tools (yt-dlp)

Reviews

No reviews yet. Be the first!

Q&A

Do I need extra tools to use the YouTube transcript feature?

The transcript extraction uses yt-dlp if it’s installed; otherwise the server falls back to extracting directly from the YouTube page, so yt-dlp is optional but recommended.

Asked by Elif Yildiz · Jan 4, 2026

What formats can fetch-mcp return for a regular webpage?

It can return raw HTML, Markdown, plain text (tags stripped), JSON (if the endpoint returns JSON), and a readable article version via Mozilla Readability, all selectable via specific commands.

Asked by Daniel Schmidt · Dec 21, 2025

How does fetch-mcp handle large responses or pagination?

It supports max_length to limit characters returned (default 5000) and start_index to begin at a specific character, allowing you to paginate through long content.

Asked by Bartek Adamski · Dec 20, 2025

Can fetch-mcp retrieve a YouTube video transcript and in which languages?

Yes, using the fetch_youtube_transcript command it can extract captions; you can specify a language code with the lang flag (default is English).

Asked by Ethan Brooks · Nov 22, 2025