PDF OCR
Turn a scanned or image-based PDF into searchable plain text — upload the PDF, get a .txt file back, with every page read by OCR on our server.
Reviewed 17 November 2026 · Free · No account needed
How to use
- Choose a PDF — scanned or image-based pages work too.
- Upload it; each page is rendered and read with OCR on our server.
- Download the plain-text result as a .txt file.
Example
Input: scanned-invoice.pdf (3 pages, no text layer)
Output: scanned-invoice.txt — the invoice text, page by page
How it works
Each page is rendered to an image at 200 DPI with pdftoppm and read by Tesseract on the server. The recognised text from all pages is joined into one plain-text file for download.
Text = Tesseract output for page 1 + page 2 + … joined with a blank line between pages. No formatting is kept — only the recognised words.
A 3-page scanned invoice becomes a .txt file with the invoice text in page order.
Edge cases to know
- Handwriting and very low-resolution scans may be recognised imperfectly.
- Pages with no readable text produce empty sections in the output.
- The output is plain text only — tables and layout are not reconstructed.
Limits & compatibility
- Handwriting and very low-resolution scans may be recognised imperfectly.
- Files up to 100 MB; the text has no formatting, only the recognised words.
- Only PDFs are accepted — image files need converting to PDF first.
Privacy
Your PDF is uploaded to our server, read with OCR, and deleted automatically after 1 hour. It is never stored, never used for anything else, and the text result is yours alone.
Guides that use PDF OCR
Related tools
Questions about PDF OCR
Is my PDF uploaded to a server?
Yes — OCR needs server processing, so the file is uploaded, read with Tesseract, and deleted automatically after 1 hour. It is never stored or used for anything else.
How accurate is the OCR?
Clean printed text at normal resolution is recognised very accurately. Handwriting, skewed pages and very low-resolution scans may have errors — always check the text before relying on it.
What do I get back?
A plain-text (.txt) file with the recognised text, page by page. Formatting, tables and images are not kept — only the words.
What is the file size limit?
100 MB. Very large PDFs take longer; each page is rendered at 200 DPI before Tesseract reads it.