OCR PDF

Recognize printed text in a PDF and download plain text. This tool does not create a searchable PDF.

Processing and network use

PDF content is processed in browser memory and is not sent to a Stocknivo conversion API. PDF.js is loaded from cdnjs; the OCR worker, runtime, and language data are served by this site, so a network connection is required and offline use is not guaranteed.

Loading OCR PDF tool...

How it works

Choose one PDF, select English, French, or Spanish, and start recognition. Pages are rendered and processed in order; the result is plain recognized text downloaded as a UTF-8 TXT file, not a searchable PDF.

Limits and recognition scope

The input is limited to 20 MiB and 25 pages. Each rendered page is limited to 4 million pixels and 4,096 pixels per canvas dimension; the document is limited to 100 million rendered pixels, 2 million recognized code points, and 10 MiB of output. Only printed English, French, and Spanish are supported. Handwriting, perfect recognition, and preservation of the original layout are not guaranteed.

Related tool

To extract text already embedded in a PDF without OCR, use PDF to TXT.