Hey! I just launched HTML to Markdown AI. It lets you drop hundreds of HTML files at once and get clean GitHub-Flavored Markdown back. No single-file copy-paste, no server uploads.
What it does:
Bulk convert entire websites to Markdown in seconds
Optimized for feeding content into LLMs and RAG pipelines
Runs 100% in your browser via WebAssembly, your files never leave your machine
Handles ZIP files with full folder structures
Why I built it:
I needed to convert hundreds of help center pages into a format I could feed into an AI knowledge base. Every tool I found either handled one file at a time or wanted me to upload everything to some server. Nothing worked at scale, so I built my own.
Why no crawling?
We deliberately don't crawl or download pages for you. There are legal grey areas around scraping other people's sites and we don't want to go there. Besides, downloading HTML is the easy part. A simple wget --mirror gets you a local copy in minutes. The actual hard part is turning that messy HTML into clean, structured Markdown. That's where this tool comes in.
Tech: Astro, Go compiled to WASM, Tailwind
Check it out: https://www.html-to-markdown-ai.com
The fact that hundreds of files stay in the browser is a really good fit for something handling private help centers and knowledge bases. The ZIP handling with folder structure is also the kind of detail you only add after actually needing it. This feels like it could be pulling in a lot more organic traffic than it probably is right now, especially with such a clear use case. Have you found that people care more about the privacy angle or the bulk conversion itself?