by puremd
Unblock, scrape, and convert web content into clean markdown while providing CDN‑style caching for AI agents, enabling reliable, rate‑limited‑free access to web pages, PDFs, images, and more.
Puremd Mcp offers tools that let AI‑driven applications retrieve web content without being blocked, transform it into pure markdown, and cache the results globally. It acts as a markdown delivery network, ensuring that repeated requests for the same resource are served quickly from cache rather than hitting the original site.
npx -y puremd-mcp.PUREMD_API_KEY for higher rate limits.unblock-url <url> – fetches the page, renders JavaScript if needed, and returns markdown.search-web <query> – performs a web search, concatenates the top results, and returns them as markdown.search-web aggregates results into a single markdown document.PUREMD_API_KEY for increased rate limits; otherwise operates anonymously.Do I need an API key? No. The service works anonymously, but providing a key raises rate limits.
What formats are supported? HTML pages, SPAs (via JS rendering), PDFs, images, and any content that can be rendered to markdown.
How is caching handled? Responses are stored globally; identical future requests are served from cache instantly.
Can I use this with other MCP clients? Yes, any client that can invoke an MCP server with a command line interface.
Is there a CLI only? The primary interface is through the two CLI commands (unblock-url and search-web), which can be invoked by any MCP‑compatible client.
Welcome to the Model Context Protocol (MCP) server for pure.md.

pure.md lets your scripts, APIs, apps, agents, etc reliably access web content in markdown format -- simply prefix any URL with pure.md/.
It avoids bot detection and renders JavaScript for SPAs, and can convert HTML, PDFs, images, and more into pure markdown. Like a CDN for markdown content, it globally caches responses for future requests to the same resource, relieving stress on origin web servers.
Without puremd-mcp, local agents may fail to fetch web content. puremd-mcp teaches MCP clients like Cursor, Windsurf, and Claude Desktop how to adopt the functionality of pure.md, giving them web unblocking and searching capabilities.
puremd-mcp comes with two tools:
unblock-url - Extract markdown from web pages without getting blockedsearch-web - Search the web for a query and concatenate results into markdownThe Model Context Protocol, developed by Anthropic, is an open standard that enables AI systems to seamlessly interact with an ecosystem of tooling. With it, MCP clients like Cursor, Windsurf, and Claude Desktop can learn how to use a variety of APIs and other functionality.
Generating an API key is an optional step that unlocks higher rate limits. If you'd like to use the pure.md MCP server anonymously, simply set your PUREMD_API_KEY value to empty string ("").
PUREMD_API_KEY value in your MCP client's configuration file (see below)Add the following to your ~/.cursor/mcp.json file:
{
"mcpServers": {
"pure.md": {
"command": "npx",
"args": ["-y", "puremd-mcp"],
"env": {
"PUREMD_API_KEY": "<TOKEN>"
}
}
}
}
Add the following to your ./codeium/windsurf/model_config.json file:
{
"mcpServers": {
"pure.md": {
"command": "npx",
"args": ["-y", "puremd-mcp"],
"env": {
"PUREMD_API_KEY": "<TOKEN>"
}
}
}
}
Add the following to your ~/Library/Application\ Support/Claude/claude_desktop_config.json file:
{
"mcpServers": {
"pure.md": {
"command": "npx",
"args": ["-y", "puremd-mcp"],
"env": {
"PUREMD_API_KEY": "<TOKEN>"
}
}
}
}
To install puremd-mcp for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @puremd/puremd-mcp --client claude
Please log in to share your review and rating for this MCP.
{
"mcpServers": {
"pure.md": {
"command": "npx",
"args": [
"-y",
"puremd-mcp"
],
"env": {
"PUREMD_API_KEY": "<YOUR_API_KEY>"
}
}
}
}claude mcp add pure.md npx -y puremd-mcpExplore related MCPs that share similar capabilities and solve comparable challenges
by modelcontextprotocol
A Model Context Protocol server that provides web content fetching capabilities.
by microsoft
Provides fast, lightweight browser automation using Playwright's accessibility tree, enabling LLMs to interact with web pages through structured snapshots instead of screenshots.
by firecrawl
Provides web scraping, crawling, search, and structured data extraction for LLM clients through a Model Context Protocol server.
by firecrawl
Provides powerful web scraping capabilities for LLM clients such as Cursor, Claude, and others, enabling content extraction, crawling, search, and batch processing through a Model Context Protocol server.
by brightdata
Provides real-time web access for AI models, delivering search, markdown‑ready scraping, and optional browser automation while bypassing geo‑restrictions, rate limits, and CAPTCHAs.
by tavily-ai
Provides real‑time web search, intelligent data extraction, structured website mapping, and systematic crawling through an MCP server, ready to be integrated with AI coding assistants and other tools.
by hyperbrowserai
Provides tools to scrape, extract structured data, crawl webpages, and access general‑purpose browser agents such as OpenAI CUA, Anthropic Claude Computer Use, and lightweight Browser Use via a Model Context Protocol server.
by zcaceres
Fetches web content and returns it in HTML, JSON, plain‑text, or Markdown formats, optionally applying custom request headers.
by tinyfish-io
Extract structured data from any web page by invoking AgentQL's extraction tool through a Model Context Protocol server, enabling AI assistants to retrieve and format web information on demand.