Website in.
LLM-ready data out.

Enter any URL. Get clean Markdown of the full site with metadata per page and structured output ready for your RAG pipeline. In one click.

webdrain
$ webdrain extract https://docs.stripe.com
Scanning sitemap...
Found ~340 pages // est. 6 min
Rendering pages...
Extracting metadata...
Cleaning content...
Done. 340 pages / 2.4 MB / 5m 42s
Saved to llms-full.txt

Extract a website

Paste any URL to scan and convert the full site to clean Markdown.

Pages
Est. time
Starting crawl... 0%
0
Pages
0
Output size
0s
Time
Download llms-full.txt

Recent crawls

No crawls yet. Paste a URL above to get started.
How it works

Three steps. Zero config.

1

Paste a URL

Enter any website. Webdrain detects sitemaps, estimates page count and crawl time before starting.

2

We crawl everything

Every subpage, every link. Metadata extracted per page. No broken half-crawls.

3

Download clean output

Get structured Markdown ready for your RAG pipeline. One file, clean format, plug and play.

Full site crawling

SPAs, docs sites, blogs — Webdrain follows sitemaps and internal links to capture every page, not just the one you pasted.

📋

Metadata per page

Title, URL, description, headings hierarchy. Every page comes with structured metadata so your pipeline knows what it's ingesting.

📊

Pre-crawl estimation

Before you start, see how many pages we'll crawl and how long it'll take. No surprises, no runaway jobs.

📁

Crawl history

All your past crawls in one place. Re-download, compare, track changes across versions. Your data pipeline's memory.

Output format

What you actually get.

llms-full.txt Ready for RAG
# Page 1 of 340
url: https://docs.stripe.com/payments
title: Accept a payment | Stripe Documentation
description: Learn how to accept a payment...
# Accept a payment
Build a payment form or use a prebuilt
checkout page to start accepting payments.
## Before you begin
This guide walks you through building a
custom payment flow using Stripe...
---
# Page 2 of 340
url: https://docs.stripe.com/webhooks
...

Stop writing scrapers.
Start shipping AI products.

Every hour you spend wrangling a crawler is an hour you're not building what matters. Webdrain exists so you never write another one.