OCR PDF: make a scanned PDF searchable
Upload a scanned PDF: every page is read with optical character recognition, and the text found is added invisibly beneath the image. You can then search for a word, select and copy, without the document looking any different.
- Free
- 15 languages
- Original image untouched
- Processed on your device
100% private — Everything is processed directly in your browser: your files, voice, and image are never sent to Chatzam's servers.
How to OCR a PDF
-
Upload the scanned PDF
The tool immediately spots which pages are images only and which already contain text.
-
Choose the languages
English is selected by default; add French, Spanish or another language if the document mixes several.
-
Download the searchable PDF
Follow the progress page by page, then get the PDF and, if you need it, the plain text as a .txt file.
A scan that behaves like a real PDF
Invisible text, word for word
Every recognized word is placed exactly over its image, so search and selection land in the right spot.
No quality loss
Page images are neither recompressed nor resized; pages that already have text are left unchanged.
Crooked pages straightened
A page scanned sideways or upside down is detected and turned the right way up before it is read.
Several languages at once
Up to 5 languages out of 15, including English, French, Spanish, German, Italian and Portuguese.
What is PDF OCR?
A PDF from a scanner or a scanning app only contains pictures of pages: you see words, but the software only sees pixels. You can't search for a name with Ctrl+F, copy a paragraph or have the document read aloud by a screen reader. OCR, short for optical character recognition, analyzes the image and finds the letters in it. This tool then places that text in the PDF as an invisible layer aligned with each word. The document looks exactly the same, but it becomes searchable, selectable and indexable, just like a PDF created straight from a word processor.
How to get the best results
Recognition quality depends first on the scan. A well-lit 300 dpi scan, in black and white or grayscale, gives excellent results on printed text. Pages are read at about 300 dpi, which leaves some margin even with a lighter scan. List every language in the document: for a bilingual English and Spanish leaflet, add both, otherwise accents and foreign words are recognized less accurately. Handwriting, stamps and text on busy backgrounds remain difficult. After OCR, run a quick test: search for an unusual word from the document to check that it's found.
Your documents stay on your device
Scanned PDFs often hold sensitive papers: IDs, pay stubs, contracts, prescriptions. Here, the file is opened and read directly in your browser; it is never sent to a server. Only the language models, a few megabytes per language, are downloaded the first time and then cached by your browser. Processing takes from a few seconds to about twenty seconds per page depending on your device; you can cancel at any time. Once the PDF is searchable, you can compress it, convert it to Word or extract its text with the other PDF tools.
Frequently Asked Questions
How do I make a scanned PDF editable?
OCR adds the recognized text to the PDF, so you can search, select and copy it. To rework the content, then convert the searchable PDF with “PDF to Word” or export the text as a .txt file.
Is my PDF uploaded to a server?
No. The PDF is read and processed entirely in your browser. Only the language models are downloaded the first time, then cached.
Is the resulting PDF larger or lower quality?
The original images are kept as they are, with no recompression. The file only grows by the size of the added text, usually a few dozen kilobytes.
What happens to pages that already contain text?
They are detected and left untouched. Only pages that are images alone go through character recognition.
Does the tool recognize handwriting?
It's designed for printed or typed text. Very neat handwriting may be partly recognized, but the result isn't reliable.