mcp-server-fetch
A Model Context Protocol server providing tools to fetch and convert web content for usage by LLMs
Decision gist · record as of 2026-08-14
Yes. The package is actively maintained, has no known vulnerabilities, installs with low friction, and fills a clear need for LLM-to-web integration in the MCP ecosystem. The permissive MIT license removes legal friction. Install it if you are running an MCP-compatible LLM platform and need web-fetching capabilities; the security caution about internal IPs is a deployment concern, not a reason to avoid the package itself.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or later.
- On Windows, may need PYTHONIOENCODING=utf-8 environment variable to avoid timeout issues.
- Optionally requires Node.js for a more robust HTML simplifier.
License · maintenance · safety
MIT (permissive) — MIT license permits free use, modification, and distribution with minimal restrictions, making it suitable for both open-source and commercial projects.
last release 2026-07-10 (35 days) · last repo commit 2026-08-10 · 89,561 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 1,086,313 downloads/mo, #4,385 on PyPI
Alternatives
Verify before relying
pip install mcp-server-fetch
python -m mcp_server_fetch
# Then configure in Claude.app or VS Code with:
# {"mcpServers": {"fetch": {"command": "python", "args": ["-m", "mcp_server_fetch"]}}}
# The server exposes a fetch tool that accepts url, max_length, start_index, and raw parameters.- Whether the server's HTML-to-markdown conversion handles complex layouts, tables, or interactive content reliably
- Performance characteristics when fetching large pages or many URLs in sequence
- How the server handles authentication-required pages or pages behind paywalls
- Specific rate-limiting or throttling behavior when used with multiple concurrent requests
What it is and what it does
mcp-server-fetch is a Model Context Protocol server that bridges LLMs and the web by providing a standardized interface to fetch and process web pages. It retrieves HTML from URLs and converts it to markdown, making web content digestible for language models. The server exposes a fetch tool with parameters for URL, maximum content length, character-index-based pagination (via start_index), and raw-content mode, allowing models to read large pages in chunks.
The package is designed to integrate with LLM platforms like Claude.app and VS Code extensions that support MCP. It respects robots.txt by default (configurable), supports custom user-agents and proxy configuration, and includes safeguards against accessing local/internal IP addresses—though the documentation cautions that this remains a potential security risk. It depends on httpx and requests for HTTP operations, markdownify and readabilipy for content extraction, and pydantic for configuration validation.
Use it for
- Enable Claude or other MCP-compatible LLMs to research topics by fetching and reading web pages during conversations
- Build automated workflows where an AI agent needs to gather information from multiple websites and synthesize it
- Integrate web-fetching capabilities into VS Code or other editors that support MCP servers for real-time web lookups
- Allow LLMs to verify claims or retrieve current information by fetching live web content
- Create chatbots or assistants that can answer questions by retrieving and processing relevant web pages on demand
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes.
The package is actively maintained, has no known vulnerabilities, installs with low friction, and fills a clear need for LLM-to-web integration in the MCP ecosystem. The permissive MIT license removes legal friction. Install it if you are running an MCP-compatible LLM platform and need web-fetching capabilities; the security caution about internal IPs is a deployment concern, not a reason to avoid the package itself.
Install
mcp-server-fetch on PyPI
Before you install
Low friction install via pip or uvx. Active maintenance with recent release (35 days old) and strong repository signals (89561 stars). Requires Python 3.10 or later. Seven runtime dependencies are all established packages.
Requires Python 3.10 or later. On Windows, may need PYTHONIOENCODING=utf-8 environment variable to avoid timeout issues. Optionally requires Node.js for a more robust HTML simplifier.
License in practice
MIT license permits free use, modification, and distribution with minimal restrictions, making it suitable for both open-source and commercial projects.
Quickstart
pip install mcp-server-fetch
python -m mcp_server_fetch
# Then configure in Claude.app or VS Code with:
# {"mcpServers": {"fetch": {"command": "python", "args": ["-m", "mcp_server_fetch"]}}}
# The server exposes a fetch tool that accepts url, max_length, start_index, and raw parameters.Verify before relying
- Whether the server's HTML-to-markdown conversion handles complex layouts, tables, or interactive content reliably
- Performance characteristics when fetching large pages or many URLs in sequence
- How the server handles authentication-required pages or pages behind paywalls
- Specific rate-limiting or throttling behavior when used with multiple concurrent requests
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 7 packageshttpxmarkdownifymcpprotegopydanticreadabilipyrequests |
| Maintenance | Actively maintained 35 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 1,086,313 / month, #4,385 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 4 - BetaIntended Audience :: DevelopersLicense :: OSI Approved :: MIT LicenseProgramming Language :: Python :: 3Programming Language :: Python :: 3.10 |
Evidence: mcp_server_fetch-2026.7.10-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “llm web content retrieval”
- mcp-server-fetchProvides a Model Context Protocol server that fetches web pages and…
- duckduckgo-mcp-serverProvides web search and content fetching through DuckDuckGo as a…
- loongsuite-util-genaiExtends OpenTelemetry's GenAI semantic conventions to capture…
Give your agent the search over MCP, or paste the wish link into any chat.
More WWW/HTTP packages
urllib3 is an HTTP client library that provides thread-safe connection pooling, SSL/TLS verification, multipart file uploads, request retries, compression support, and proxy handling for Python applications.
Requests is a Python HTTP library that simplifies sending HTTP/1.1 requests with automatic handling of headers, authentication, cookies, and response parsing.
h11 is a pure-Python HTTP/1.1 protocol implementation that handles parsing and serializing HTTP messages without any built-in I/O, letting you integrate it with any network layer you choose.
HTTPX is a fully featured HTTP client library for Python that provides both sync and async APIs, with support for HTTP/1.1 and HTTP/2, plus an integrated command-line client.
Install it if you are building new projects or modernizing existing ones that rely on HTTP.
A minimal low-level HTTP client library that sends HTTP requests with thread-safe and task-safe connection pooling, supporting HTTP/1.1, HTTP/2, proxies, and both sync and async interfaces.
aiohttp is an async HTTP client and server framework built on asyncio, supporting both WebSockets and middleware-based routing for building concurrent web applications.
Install it if you need async HTTP client or server capabilities in asyncio-based applications.
See also mcp-server-git · mcp-server-time · mcp · langsmith-mcp-server · arxiv-mcp-server · cli-mcp-server · duckduckgo-mcp-server · web-forager · markitdown-mcp · duckduckgo-mcp