Guide · Converting
How to OCR a PDF
Reviewed
Short answer
To OCR a PDF, upload the scanned file to an OCR tool, choose the language printed on the pages, and run recognition. The result looks the same but has an invisible text layer, so you can search with Ctrl+F (Cmd+F on a Mac), copy text and use AI tools on it. Always test by searching for a few words, because recognition errors don't show on the page.
On this page (7)
Does your PDF need OCR?
Press Ctrl+F (Cmd+F on a Mac) and search for a word you can see. If nothing is found, or you can't select a single word with the cursor, the page is an image and needs OCR. Files from scanners, phone scanning apps and fax services are the usual cases.
1A scan: only pixels
2Recognize the text
3Text you can search and select
OCR a PDF in PDFBetter
- 1
Upload the scan
Open OCR PDF and add a PDF, or a JPG, PNG or TIFF image.
- 2
Choose the languages
Select every language that appears in the document. Accents and non-Latin scripts depend on this.
- 3
Run OCR
Recognition runs on PDFBetter's processing servers. Each page counts toward your monthly OCR allowance.
- 4
Check and download
Search for a word you can see on the page to confirm it was recognized, then download the searchable PDF.
Get better recognition
Scan at 300 dpi; lower resolutions lose small characters and higher ones rarely help. Keep pages straight and well lit, and avoid shadows from phone photos. If a scan is crooked, faint or noisy, clean it up before OCR. Clean printed text recognizes well; handwriting, stamps over text and decorative fonts don't.
What OCR does and doesn't give you
OCR adds searchable, copyable text behind the image. It doesn't rebuild headings, columns or tables, and the visible page doesn't change. To change words on the page, use the editor after OCR. To get an editable document, convert the OCR'd PDF to Word.
Reading order can be wrong on multi-column pages and forms, which you'll notice when copying long passages.
When to use AI OCR
For dense tables, handwriting and poor scans, AI OCR adds a vision-model correction pass over standard recognition. It costs AI credits per page, so use it on the pages that need it rather than the whole document.
Tools used in this guide
- OCR PDFUpload a scan, pick its language, and download a PDF that looks the same but lets you search, select and copy the text.
- AI OCRRun OCR, let a vision model correct what it misread, and download a searchable PDF with better recognized text.
- Restore scanned PDFClean up crooked, grey or speckled scan pages and download a restored copy, keeping the original in case you prefer it.
- Edit scanned PDFUpload a scan, run OCR in your document's language, and keep working in the editor with text you can select, correct and add to.
- Search PDFType a word or phrase and get every match in the PDF, listed by page with surrounding text, ready to click through.
Questions people ask
How many pages can I OCR for free?
5 pages without an account, 20 a month on the Free plan and 1,000 a month on Pro.
Does OCR change the look of my document?
No. The page image stays exactly as it was; only an invisible text layer is added.
Can OCR read handwriting?
Standard OCR recognizes handwriting poorly. AI OCR handles neat handwriting better, but check every word.
Can I OCR a password-protected PDF?
Unlock it with its password first.
Why does copied text come out jumbled?
On multi-column layouts, OCR can read across columns instead of down them. Copy one column at a time or convert to Word and fix the order there.