Web2MD: URL to clean Markdown

Turn any URL into clean Markdown that AI agents can actually use

Problem

When feeding web pages to AI agents, copying raw HTML brings in navigation, ads and scripts that waste context and hurt comprehension. The Show HN pitches 'turn any URL into clean Markdown for AI agents,' confirming this is a recurring friction in agent workflows.

Solution

A thin, focused converter: paste a URL, extract the main article, strip ads and nav, preserve heading hierarchy, code blocks and tables, and output clean Markdown with optional link footnotes. Ship as a browser extension, CLI and API so it can plug into agent pipelines.

Users

AI agent developers, RAG engineers, researchers and writers who assemble web material into knowledge bases

MVP

Just URL in, Markdown out; article extraction with noise stripping; preserve headings, lists and code blocks; one-click browser extension copy; CLI that writes to clipboard; no batch crawling in v1

Revenue

Free CLI and extension; usage-based API pricing; Pro tier for batch jobs and more aggressive cleanup; or a one-time desktop purchase

Why now

Agents and RAG are now part of everyday dev workflows, but clean web input remains painful. Existing tools are either too heavy or too expensive, leaving room for a small, focused converter