DocStripper automatically removes noise from text documents: page numbers, headers, footers, duplicate lines, and empty lines. Choose between Fast Clean (instant rule-based) or Smart Clean (AI-powered with on-device LLM). Works entirely in your browser - no uploads, no sign-ups, 100% private. Upload a document, configure cleaning options, and download clean results instantly. Perfect for students, researchers, and anyone working with scanned documents or PDFs. Also available as a free CLI tool.

Hey Product Hunt! 👋

I'm excited to launch DocStripper today — a tool I built to solve my own problem of cleaning messy documents.

The Problem: I was spending hours manually removing page numbers, headers, and duplicate lines from scanned documents and PDFs. There had to be a better way.

The Solution: DocStripper cleans documents automatically, right in your browser. No uploads, no sign-ups, completely private.

Why I built it:

- 100% client-side processing (your files never leave your device)

- AI-powered Smart Clean mode — Choose between Fast Clean (instant rule-based) or Smart Clean (on-device LLM) for intelligent cleaning

- Works offline after first load

- Free and open-source

- Both web app and CLI tool available

Try it out: Upload a document at https://kiku-jw.github.io/DocStripper/ and see the magic happen! You can test both Fast Clean and Smart Clean modes — the AI mode runs entirely in your browser using WebGPU.

I'd love to hear your feedback and use cases. What types of documents do you clean most often? Any features you'd like to see added?

Thanks for checking it out! 🚀

Hmm, this is intriguing. @kiku_reise I faced similar problems.
Damn! I never imagined that such a tool would be available.
Brilliant idea