A flexible HTTP fetching Model Context Protocol server.
A flexible HTTP fetching Model Context Protocol server. That is what the fetch mcp server brings to an AI assistant: the same capability, reachable through the Model Context Protocol rather than a separate app or dashboard.
An MCP server for fetching web content in multiple formats — HTML, JSON, plain text, Markdown, readable article content, and YouTube transcripts.
Installation goes through your MCP client rather than a global install: point it at mcp-fetch on npm and it is fetched when the client starts. The copy-paste blocks for Claude Desktop, Claude Code and Cursor are further down this page.
The server publishes 14 tools. What each one is for:
fetch_html — Fetch a website and return its raw HTML contentfetch_markdown — Fetch a website and return its content converted to Markdownfetch_txt — Fetch a website and return plain text with HTML tags, scripts, and styles removedfetch_json — Fetch a URL and return the JSON responsefetch_readable — Fetch a website and extract the main article content using Mozilla Readability, returned as Markdownfetch_youtube_transcript — Fetch a YouTube video's captions/transcript. Uses yt-dlp if available, otherwise extracts directly from the page. Accepts an additional langurl — stringheaders — objectmax_length — numberstart_index — numberproxy — stringhtml — Fetch a URL and return raw HTMLConfiguration is passed through the environment: DEFAULT_LIMIT. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.
Plenty of search and retrieval servers cover similar ground. The differences that matter in practice are scope of access and how much setup stands between you and a working tool call. Fetch's toolset — fetch_html, fetch_markdown, fetch_txt and 11 more — is a fair guide to whether it matches your workflow. It is maintained by zcaceres; worth a glance at recent repository activity before you build anything load-bearing on it.
SyncDev reviews every entry in this directory against the project's own documentation before publishing, and revisits them as servers change.
| Tool | What it does |
|---|---|
| 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](https://github.com/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. |
| url | string |
| headers | object |
| max_length | number |
| start_index | number |
| proxy | string |
| 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) |
{
"mcpServers": {
"fetch": {
"command": "npx",
"args": ["mcp-fetch-server"]
}
}
}Configuration as documented by the project. Restart the client after saving.
| Variable | Description | Required |
|---|---|---|
| DEFAULT_LIMIT | Configuration value read at startup. | Optional |
The simplest web tool that matters — fetch any URL and get model-ready markdown back.
Industrial-strength web extraction — render, scrape, crawl and search entire sites into clean markdown.
Puppeteer-powered browser control that drives pages from the accessibility tree instead of pixels.
Search built for AI, not humans — semantic web search that returns model-ready content, plus code context.
Answers, not links — delegate questions to Perplexity's search-grounded models and get cited responses back.
Put 6,000+ pre-built scrapers at your assistant's fingertips through one MCP endpoint.