You scanned a contract or received a photographed document, and now you can't search it, copy a sentence or paste a figure into an email. That's because the file contains a picture of text, not the text itself. OCR, short for optical character recognition, reads the picture and recovers the words.
How to tell if you need OCR
Open the PDF and press Ctrl+F (Cmd+F on Mac) to search for a word you can see on the page. If nothing is found, or you can't highlight individual words, the page is an image.
How to run OCR
- Open OCR PDF.
- Upload your scanned file (up to 50 MB).
- Start the process and wait for it to finish.
- Download the result and check it.
The OCR tool is tuned for English text at the moment, so expect the best results with English documents.
What affects accuracy
OCR quality depends mostly on the scan, not the software.
- Resolution. 300 DPI is a good target for text. Far below 150 DPI and small letters blur into each other.
- Contrast. Dark text on a light background reads best. Faded photocopies and shadows cause errors.
- Straightness. A skewed page confuses line detection. Fix sideways or upside-down pages with Rotate PDF before running OCR.
- Layout. Simple single-column pages work best. Tables, columns and handwriting are harder, and handwriting is generally not reliable.
Always proofread
Even good OCR makes mistakes, and they cluster in the places you care about most: numbers, names and dates. Confusing "0" and "O", or "1" and "l", is classic. Read through anything you'll rely on, and check figures against the original image.
What to do with the text
- Copy it straight into an email or document, or use Extract Text to pull it out in bulk.
- Edit it by converting to Word with PDF to Word. See how to convert without losing formatting.
- Move tables into a spreadsheet with PDF to Excel, then check the numbers carefully.
- Keep a searchable archive. A searchable PDF is far easier to find things in months later.
A five-minute proofreading routine
Search the result for the words you care about most, such as names and reference numbers. Check every number and date against the original page image, because that is where mistakes matter most. Look for common mix-ups: the letter O and the digit 0, lowercase l and the digit 1, and rn read as m. Read one full paragraph aloud against the page. If errors keep appearing, rescan at a higher resolution or with better lighting and run OCR again instead of fixing everything by hand.
Frequently Asked Questions
Does OCR change how the document looks?
The page looks the same. OCR adds a text layer that you can search and select.
Can it read handwriting?
Not reliably. It is built for printed text.
Why are some words wrong?
Usually low resolution, poor contrast or an unusual font. A cleaner, straighter scan almost always improves the result.
Will OCR make my file bigger?
Sometimes slightly. If it grows too large, compress it afterwards with Compress PDF.