OCR PDF
Run text recognition on a scanned or image-based PDF (photos of documents, scanned forms) to pull out real, selectable, searchable text.
OCR runs entirely in your browser using an on-device recognition model (nothing is uploaded), which means it can take a little while for longer documents — a progress bar will show you how far along it is. Larger files show a short ad before the text is unlocked.
About OCR accuracy
How accurate is it?
Accuracy depends heavily on scan quality — clean, high-resolution, straight scans give the best results. Handwriting and very low-quality photos may not recognize well.
Why does it take time?
The recognition model runs fully on your device instead of a server, which keeps your document private but means larger or multi-page PDFs take longer than a cloud OCR service.
Is my file uploaded anywhere?
No — OCR happens entirely in your browser. Your PDF never leaves your device.