OCR PDF
Make scanned PDFs searchable and selectable with local text recognition.
Free · No upload · 100% on your device
Make a scanned PDF searchable
Choose a PDF, or drop it here.
No file-size or page limits. Nothing is uploaded.
You chose more than one PDF. What would you like to do?
100% on your device
How to use OCR PDF
- Choose your PDF, or drop it on the page.
- Adjust the settings in the side panel. The preview updates as you go.
- Select “Make searchable”.
- Download the result. Your original file is left unchanged.
Private by design. Your files are processed in this browser tab and are never uploaded to KittySuite.
About OCR PDF
Optical character recognition reads the words printed on a scanned or photographed page and stores them as an invisible text layer behind the image. The page looks the same, but you can now search it with Ctrl+F, select and copy sentences, and let screen readers and document search index it. Choose the document language first for the best accuracy.
Pages that already contain real text are skipped by default, so a mixed PDF keeps its original text untouched. Accuracy depends on the scan: sharp, straight, well-lit pages at 300 dpi read best, while handwriting, tiny print, heavy stamps and unusual fonts produce more mistakes. Check important numbers against the page image. Recognition runs with Tesseract in this browser tab, typically a few seconds per page.
OCR PDF: frequently asked questions
How do I make a scanned PDF searchable?
Choose the PDF and its language, then run OCR. The recognised words are added as an invisible layer behind each page, so the page looks the same but can be searched and copied.
Which languages are supported?
English, Spanish, French, German, Portuguese and Italian.
How accurate is the text recognition?
Sharp, straight, well-lit pages at about 300 dpi read best. Handwriting, tiny print, stamps and unusual fonts produce more mistakes, so check important numbers.
Are pages that already contain text processed again?
Not by default; they are skipped so their original text stays untouched. You can also save the recognised text as a .txt file.
Does OCR run on a server?
No. The OCR engine and language model, about 3 MB each, download once from this site and run in your browser, usually a few seconds per page.