PDF OCR: turn image into editable text
NoCatchPDF runs OCR (text recognition) right in your browser, in Portuguese and English. Choose to recognize the whole document or just a selected area: recognized text becomes editable text boxes over the document. Nothing is sent to servers — neither the PDF nor the text.
Recognize text now →How it works
- Open the editor. Click the button below and open the scanned PDF.
- Load OCR. The first time, click load text recognition to download the model (~19 MB, saved for offline use).
- Choose the scope. Recognize the whole document or draw an area to recognize just that part.
- Edit the result. Recognized text appears in editable boxes: fix, merge boxes, and adjust font and size.
Why is it free and unlimited?
Because processing happens in your browser, not on our servers. We don't pay for expensive infrastructure to process your files — so we don't need to charge you, make you sign up, or cap your usage. No catch: that's literally what the name means.
FAQ
Which languages does the OCR recognize?
Portuguese and English, with professional quality (the Tesseract engine, the same used in archive digitization projects).
Why do I need to download a model the first time?
The OCR engine is ~19 MB and runs entirely on your computer. Instead of making the site heavy for everyone, it's downloaded only when you ask — and stays saved to work offline afterwards.
Does OCR make the PDF searchable?
Recognized text is inserted as editable text boxes over the document, which you can review before exporting. A searchable-PDF mode (invisible text layer) is on our roadmap.
Does the recognized text go to the internet?
No. Recognition happens 100% in your browser; neither the PDF nor the text leaves your computer.