OCR a PDF, free, in your browser
Make a scanned PDF searchable: select, copy and find its words. The text is read on your device and laid invisibly over the scan, so every page looks exactly as before.
✓ no upload✓ pages unchanged✓ 11 languages✓ works offline
How OCR works here
- Drop the scan. A PDF from a scanner, a phone app or a fax. Pick the language it is written in; a second one if the pages mix two.
- Read. Each page without text is drawn at 300 dpi and read by Tesseract 5, the open source OCR engine, running in your browser. The first time, the reader and the language are fetched from this site and kept.
- Download. Every word goes back onto the page, invisible, exactly where the scan shows it. The result card lists the pages read and how sure the reader was.
What it does
- Reads the text on your device: a scanned contract, a passport or a medical letter never leaves it.
- Keeps the page as it was, pixel for pixel; the file grows by the text alone.
- Places each word over its picture, so selecting and highlighting line up with what you see.
- Turns along with a page scanned sideways and rotated upright.
- Says how sure it is on average, warns for each page under 60 %, and names the words it was least sure of.
What it does not do
- OCR guesses: small print, poor scans and unusual fonts come out with mistakes, and the result does not hide that.
- It does not read handwriting reliably.
- It does not straighten a crooked scan or clean up the picture; it reads the page as it is.
- It does not make the text editable in the page itself; PDF to Word makes an editable copy.
Questions
What does OCR do to my PDF?
It adds the text it reads as an invisible layer over each scanned page. The scan stays exactly as it was. Afterwards you can search the file, select words and copy them.
How accurate is it?
On a clean scan of printed text at 300 dpi, almost every word comes out right. Blurry photos, tiny print and handwriting do worse. The result gives the average confidence, names every page under 60 % and the words it was least sure of, so you know where to check.
Which languages can it read?
English, German, French, Spanish, Italian, Portuguese, Dutch, Polish, Czech, Croatian and Russian. Pick two at once for a document that mixes them; it is slower.
Why does the first run take longer?
Your browser fetches the reader, about 1.5 MB, and the language, 0.7 to 3 MB, from this site once and keeps them. After that, a full page takes a few seconds.
Does my file leave my computer?
No. The reading runs in your browser tab. The only network requests are the site's own files, which you can check by switching the connection off after the first run.
Guides for this tool
- Make a scanned PDF searchable, without changing how it looksOCR a scanned PDF on your device: free, no account, nothing is uploaded.
- Bank statement to Excel, without uploading it anywhereTurn a bank statement PDF into an Excel sheet without uploading it anywhere: steps, amount checks, and what to do when the statement is a scanned image.
- Convert many PDFs to text at once, and catch the scans that come out emptyConvert multiple PDFs to text in one batch in your browser: a whole folder in, 1 .txt per PDF out, and how to spot the scanned files that come out empty.
- Cannot select or copy text in a PDF: the 3 causes and their fixesText in a PDF will not select or copy for 1 of 3 reasons: a scan, a copy restriction, or a font that hides the letters.
- Text copied from a PDF comes out jumbled: why, and what fixes itCopy and paste from a PDF comes out jumbled: wrong order, gibberish or glued words.
- Edit a scanned PDF: change the words on a scan, then make it searchableA scanned PDF is a photo of text, so nothing selects.