Reader Markdown
Clean markdown from any URL.
Just add the prefix.
https://md.reader.dev/https://example.comRaw HTML is noisy. Full of scripts, nav bars, ads, and markup that confuses language models. Reader strips it all away and returns clean, structured markdown that LLMs can actually understand.
No API key needed. No setup. No code.
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width">
<title>About Our Company</title>
<link rel="stylesheet" href="/assets/main.css">
<script src="/analytics.js"></script>
<script src="/tracking.min.js"></script>
</head>
<body>
<nav class="nav-wrapper bg-white shadow-sm">
<div class="container mx-auto flex items-center">
<a href="/" class="logo"><img src="/logo.png"></a>
<ul class="nav-links">
<li><a href="/products">Products</a></li>
<li><a href="/pricing">Pricing</a></li>
<li><a href="/about">About</a></li>
<li><a href="/contact">Contact</a></li>
</ul>
</div>
</nav>
<main class="container mx-auto py-12">
<div class="hero-section">
<h1 class="text-4xl font-bold mb-4">
About Our Company
</h1>
<p class="text-lg text-gray-600 mb-8">
We build tools that help teams work faster.
Founded in 2020, we serve over 10,000
customers worldwide.
</p>
<h2 class="text-2xl font-semibold mb-3">
Our Mission
</h2>
<p class="text-base text-gray-500">
Making complex workflows simple through
intelligent automation and thoughtful
design.
</p>
</div>
</main>
<div id="cookie-banner" class="fixed bottom-0">
<p>We use cookies...</p>
</div>
<script>trackPageView('about');</script>
</body>
</html># About Our Company
We build tools that help teams work faster.
Founded in 2020, we serve over 10,000 customers
worldwide.
## Our Mission
Making complex workflows simple through
intelligent automation and thoughtful design.Same page. One line of code. Completely different input quality.
Why markdown matters for LLMs
HTML wastes tokens. A typical webpage is 80% markup and 20% content. Navigation menus, script tags, style blocks, tracking pixels, cookie banners. Your model processes all of it and extracts meaning from almost none of it.
Markdown is pure signal. Headings convey structure. Lists convey hierarchy. Links are preserved with context. Nothing is wasted. Every token carries meaning.
Reader converts any public webpage into markdown that is optimized for language models. Fewer tokens, better comprehension, more accurate outputs.
How it works
Three steps. That's it.
Prefix any URL
https://md.reader.dev/ + the page you want to read.
Reader fetches, renders, and cleans
JavaScript execution, content extraction, boilerplate removal. All handled automatically.
Get clean markdown back
Ready to feed into your RAG pipeline, agent workflow, prompt chain, or just read yourself.
Powered by Supermarkdown
The conversion engine behind Reader is Supermarkdown, our HTML-to-markdown library built from scratch in Rust. We didn't wrap an existing converter or shell out to a Node.js library. We wrote a purpose-built parser because timing is everything when you're converting millions of pages.
Single-pass O(n) conversion. Sub-5ms for a typical webpage. 10 to 50x faster than JavaScript alternatives. Because when your agent is waiting on web data, every millisecond counts.
Use cases
RAG pipelines
Feed clean web content into your retrieval system without preprocessing.
AI agents
Give your agents the ability to read any webpage as part of their tool chain.
Research
Pull content from sources without copy-pasting or fighting with HTML parsers.
Documentation
Convert any docs site into markdown for local reference or fine-tuning datasets.