A scanned PDF is a picture. Search returns nothing. Copy-paste returns nothing. The text exists visually but the computer can't read it.
OCR (optical character recognition) reads the visible text and adds a hidden text layer behind it. The scan looks the same; the document becomes fully searchable.
When OCR is needed
If you can't select text in your PDF, it needs OCR. Common cases: scanned contracts, photographed documents, screenshots saved as PDF.
Live-text PDFs (exported from Word, Pages, etc.) don't need OCR — the text is already there.
Run OCR in the editor
Open the PDF in Flint's editor and run OCR. The tool processes every page and adds a recognised text layer.
Specify the language for best accuracy. English defaults to fast and accurate; other languages slightly slower with similar accuracy.
Verify and clean up
Try searching for a word you know is in the document. If results land on the right page, OCR worked.
For key fields (numbers, names, dates), spot-check the recognised text for accuracy. OCR isn't perfect — 90-95% accurate on clean scans, less on messy ones.
FAQ
How accurate is OCR?
90-95% for clean printed text. 70-85% for low-contrast or skewed scans. 40-70% for handwriting.
Will OCR change how the PDF looks?
No. The visible scan is unchanged. The text layer is invisible and sits behind the scan.
Can I OCR a multi-language PDF?
Yes — specify multiple languages. Accuracy drops slightly versus single-language.
Does OCR work on photos of documents?
Yes. Convert the photo to PDF first, then OCR. Cleaner photos give better results.
OCR is the difference between a picture and a document. Run it in Flint's editor and your scan becomes a real searchable file.