Convert any URL to clean Markdown — directly from your AI agent.
An MCP (Model Context Protocol) server that converts public webpages to clean, readable Markdown. Strips ads, navigation, and boilerplate — returns LLM-ready text perfect for RAG, research, and content extraction.
Works with Claude Desktop, Cursor, Continue.dev, Windsurf, and any MCP-compatible client.
git clone https://github.com/astra-intelligence/web2md-mcp.git
cd web2md-mcp
pip install .
⏳ Pending PyPI publication. Until then, use the source install above.
pip install web2md-mcp
Add to your Claude Desktop config (claude_desktop_config.json):
{
"mcpServers": {
"web2md": {
"command": "uvx",
"args": ["web2md-mcp"]
}
}
}
The server provides one tool: web2md_convert
| Parameter | Type | Description |
|---|---|---|
url |
string | The public URL to convert to Markdown |
{
"url": "https://example.com/article"
}
Returns clean Markdown content ready for LLM consumption.
- Python 3.10+
- Internet access to reach the Web2MD API at
167.233.135.161:9999
Web2MD is a URL-to-Markdown conversion API running at 167.233.135.161:9999. This MCP server wraps the API so AI agents can use it directly.
Free tier: 10 conversions/day per IP — no signup, no API key.
👉 Get unlimited access on Gumroad: https://grantshatz.gumroad.com/l/mpkqyq ($1+)
You can also use the API directly without the MCP server:
curl "http://167.233.135.161:9999/api/convert?url=https://example.com/"
Free tier: 10/day/IP. Add ?license_key=YOUR_KEY for unlimited access.
MIT
If you find this project useful, consider buying me a coffee or sponsoring on GitHub.
Check out more tools at grantshatz.gumroad.com.
Generate a free GitHub profile stats card (stars, languages, followers) for your README: GitHub Stats Card