Never paste a messy PDF into ChatGPT again
Drop your file here. Get clean Markdown that AI understands perfectly. Word count, tokens, download — all processed locally in NL.
Why KlareTekst?
Clean Markdown
No weird characters or layout artifacts. Drop into ChatGPT, Claude, Gemini or your RAG pipeline instantly.
100% privacy-first
Files processed in memory and deleted immediately. No storage, no tracking. Server in NL.
Built-in token counter
See immediately if output fits your AI model — no surprises at prompt limits.
Batch + OCR (Pro)
Pro users: up to 100 files/day, 25 MB each, scanned PDFs via OCR, ZIP download.
Which formats are supported?
Drop multiple files at once — all get converted one by one.
- PDFPDF — text + tables
- DOCXDOCX — Word documents
- XLSXXLSX — Excel sheets (as Markdown tables)
- CSVCSV — comma-separated data
- HTMLHTML — web pages
- TXTTXT — plain text
- MDMD — Markdown (validate + clean)
- OCRPNG/JPG — Pro (OCR via Tesseract)
How does it work?
1. Drop file
Drop your PDF, Word or Excel. Multiple at once works too.
2. We extract
Server processes locally in NL, deletes immediately. Result = clean Markdown.
3. Copy or download
Paste straight into ChatGPT/Claude, or download as .md / .zip.
Pro plan — €9/month
For consultants, lawyers and researchers working with AI daily.
- 100 files/day · 25 MB each
- OCR for scanned PDFs and PNG/JPG
- Batch download as ZIP
- Writer-focus mode (no images)
- Priority support + API access
FAQ
Do you store my documents?
No. Files are processed in RAM and deleted right after conversion. We only log an anonymous counter (conversions per IP-hash for rate limiting).
Is it GDPR-compliant?
Yes. Server in NL (Hetzner), no third-party conversion APIs, no cookies except "lang" preference, no tracking. GDPR-ready for sensitive docs.
Why Markdown?
LLMs read Markdown much better than raw PDF text. Headings, lists and tables stay intact, which dramatically improves prompt quality.
Does it work for scanned PDFs?
Only with Pro (Tesseract OCR). Free plan works with text PDFs.