What formats can the Fetch MCP Server retrieve web content in?

The Fetch MCP Server can retrieve content in six formats: HTML (raw), JSON (API responses), plain text (with tags and scripts stripped), Markdown (converted content), readable article content (extracted via Mozilla Readability), and YouTube transcripts (via captions).

Does the Fetch MCP Server work with sites that require authentication?

It supports custom headers, which can include authentication tokens, but it is not designed for sites requiring complex authentication beyond custom headers.

How do I use the Fetch MCP Server as a command-line tool?

You can run it as a standalone CLI using `mcp-fetch [flags]`, supporting the same six fetch operations plus flags like `--max-length`, `--start-index`, `--proxy`, and `--lang`.

Can the Fetch MCP Server extract content from JavaScript-rendered pages?

No, it fetches and converts static or server-rendered content only—it is not a headless-browser rendering engine and cannot handle JavaScript-rendered content.

MCP Connector

Fetch Web Content in Multiple Formats

Name: Fetch MCP Server
Availability: OnlineOnly
Author: Zach Caceres

The Fetch MCP server retrieves web content as HTML, Markdown, JSON, readable article text, or YouTube transcripts.

Connect

Zach Caceres

Own this? Claim it

Maintainer of this project? Claim this page to edit the listing.

Spark score

out of 100

Updated 4 months ago

Version 1.0.0

Models

universal

Add to Favorites

Why it matters

This asset acts as a server to fetch web content, providing it in various formats like HTML, JSON, plain text, and Markdown. It supports custom headers and pagination for flexible data retrieval.

Outcomes

What it gets done

Fetch web pages as HTML

Extract JSON data from URLs

Retrieve plain text content

Convert HTML to Markdown

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/vb-fetch | bash

Capabilities

Tools your agent gets

fetch_html

Fetches a website and returns the content in HTML format

fetch_json

Fetches a JSON file from a URL

fetch_txt

Fetches a website and returns the content as plain text without HTML

fetch_markdown

Fetches a website and returns the content in Markdown format

Overview

Fetch MCP Server

The Fetch MCP server retrieves web content as HTML, Markdown, plain text, JSON, readable article content via Mozilla Readability, or YouTube transcripts, with built-in SSRF protection. Use it when Claude needs to fetch and read web content in a specific clean format, including readable article extraction or YouTube transcripts.

What it does

The Fetch MCP server fetches web content in multiple formats: HTML, JSON, plain text, Markdown, readable article content, and YouTube transcripts. It exposes six tools sharing common parameters (URL, custom headers, max character length, start index for pagination, and an optional proxy): fetch_html (raw HTML), fetch_markdown (content converted to Markdown), fetch_txt (plain text with tags, scripts, and styles stripped), fetch_json (the JSON response from an API URL), fetch_readable (main article content extracted via Mozilla Readability, stripped of navigation, ads, and boilerplate, returned as Markdown - ideal for articles and blog posts), and fetch_youtube_transcript (a video's captions via yt-dlp or direct page extraction, with a lang parameter defaulting to English).

It can run as an MCP server or as a standalone CLI (mcp-fetch <command> <url> [flags]) supporting the same six operations plus --max-length, --start-index, --proxy, and --lang flags. Two environment variables tune server-wide behavior: DEFAULT_LIMIT (default character limit for responses, 5000, or 0 for no limit) and MAX_RESPONSE_BYTES (maximum response body size, default 10 MB). Built-in protections include SSRF protection (blocking private/localhost addresses and DNS rebinding) and response size limits to prevent memory exhaustion.

npm install -g mcp-fetch-server

When to use - and when NOT to

Use this connector when Claude needs to fetch and read web content in a specific format - extracting clean article text from a blog post, converting a page to Markdown, pulling a YouTube transcript, or fetching a JSON API response - especially through a proxy or with pagination for long pages.

It is not a fit for sites requiring authentication beyond custom headers, or as a general-purpose scraper for JavaScript-rendered content - it fetches and converts static or server-rendered content in the formats listed, not a headless-browser rendering engine.

Capabilities

Six fetch tools sharing URL, headers, max-length, start-index, and proxy parameters: fetch_html, fetch_markdown, fetch_txt, fetch_json, fetch_readable (Mozilla Readability article extraction), and fetch_youtube_transcript (with language selection). Built-in SSRF protection and response-size limits guard against private-network access and memory exhaustion.

How to install

Install globally with the command above, or run directly via npx mcp-fetch-server as an MCP server, or npx mcp-fetch <command> <url> [flags] as a standalone CLI. Add it to your MCP client configuration with npx and mcp-fetch-server as the command and args, optionally setting DEFAULT_LIMIT or MAX_RESPONSE_BYTES as environment variables. The project is MIT-licensed.

Who it's for

Anyone who wants Claude to fetch and read web content in a clean, specific format - readable article text for research, Markdown for documentation, JSON for API responses, or YouTube transcripts - without manually copying and pasting page content into the conversation.

Source README

Fetch MCP Server

An MCP server for fetching web content in multiple formats - HTML, JSON, plain text, Markdown, readable article content, and YouTube transcripts.

Tools

All tools accept the following common parameters:

Parameter	Type	Required	Description
`url`	string	Yes	URL to fetch
`headers`	object	No	Custom headers to include in the request
`max_length`	number	No	Maximum characters to return (default: 5000)
`start_index`	number	No	Start from this character index (default: 0)
`proxy`	string	No	Proxy URL (e.g. `http://proxy:8080`)

fetch_html - Fetch a website and return its raw HTML content.
fetch_markdown - Fetch a website and return its content converted to Markdown.
fetch_txt - Fetch a website and return plain text with HTML tags, scripts, and styles removed.
fetch_json - Fetch a URL and return the JSON response.
fetch_readable - Fetch a website and extract the main article content using Mozilla Readability, returned as Markdown. Strips navigation, ads, and boilerplate. Ideal for articles and blog posts.
fetch_youtube_transcript - Fetch a YouTube video's captions/transcript. Uses yt-dlp if available, otherwise extracts directly from the page. Accepts an additional lang parameter (default: "en") to select the caption language.

Installation

As an MCP server

Add to your MCP client configuration:

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

As a CLI

npx mcp-fetch <command> <url> [flags]

Or install globally:

npm install -g mcp-fetch-server
mcp-fetch <command> <url> [flags]

CLI Usage

mcp-fetch <command> <url> [flags]

Commands

Command	Description
`html`	Fetch a URL and return raw HTML
`markdown`	Fetch a URL and return Markdown
`readable`	Fetch a URL and return article content as Markdown (via Readability)
`txt`	Fetch a URL and return plain text
`json`	Fetch a URL and return JSON
`youtube`	Fetch a YouTube video transcript

Flags

Flag	Description
`--max-length <N>`	Maximum characters to return
`--start-index <N>`	Start from this character index
`--proxy <URL>`	Proxy URL
`--lang <code>`	Language code for YouTube transcripts (default: `en`)
`--help`	Show help message
`--version`	Show version

Examples

# Fetch a page as markdown
mcp-fetch markdown https://example.com

# Extract article content without boilerplate
mcp-fetch readable https://example.com/blog/post

# Get a YouTube transcript in Spanish
mcp-fetch youtube https://www.youtube.com/watch?v=dQw4w9WgXcQ --lang es

# Fetch with a length limit
mcp-fetch html https://example.com --max-length 10000

# Fetch through a proxy
mcp-fetch json https://api.example.com/data --proxy http://proxy:8080

Environment Variables

Variable	Description
`DEFAULT_LIMIT`	Default character limit for responses (default: `5000`, set to `0` for no limit)
`MAX_RESPONSE_BYTES`	Maximum response body size in bytes (default: `10485760` / 10 MB)

Example with a custom limit:

{
  "mcpServers": {
    "fetch": {
      "command": "npx",
      "args": ["mcp-fetch-server"],
      "env": {
        "DEFAULT_LIMIT": "50000"
      }
    }
  }
}

Features

Fetch web content as HTML, JSON, plain text, or Markdown
Extract article content with Mozilla Readability (strips ads, nav, boilerplate)
Extract YouTube video transcripts (via yt-dlp or direct extraction)
Proxy support for requests behind firewalls
Pagination with max_length and start_index
Custom request headers
SSRF protection (blocks private/localhost addresses and DNS rebinding)
Response size limits to prevent memory exhaustion

Development

bun install
bun run dev     # start with watch mode
bun test        # run tests
bun run build   # build for production

FAQ

Common questions

Discussion

Fetch Web Content in Multiple Formats

What it gets done

Add it to your toolbox

Tools your agent gets

Fetch MCP Server

What it does

When to use - and when NOT to

Capabilities

How to install

Who it's for

Fetch MCP Server

Tools

Installation

As an MCP server

As a CLI

CLI Usage

Commands

Flags

Examples

Environment Variables

Features

Development

Common questions

Questions & comments · 0