How to convert a scanned PDF to Word
Updated
Contents
To turn a scanned PDF into a Word document you can edit, open PDF to Word, add the PDF, tick “Recognise text in scans (OCR)”, pick the languages the document is written in under Languages, and press Export DOCX. Our server reads the words on each scanned page with OCR (optical character recognition) and writes them into a DOCX file as text you can edit, copy and search.
A scanned page holds a picture of its words, not the words themselves, so without that switch there is no text on it for the conversion to carry. With it, the panel says what happens: “Runs OCR on our secure server first, in the languages you pick, then converts, all in one job. Scanned pages keep their picture and gain their words as text; pages that already have text are kept as they are.”
This runs on our secure server, not in your browser. Our privacy policy says the uploaded file is deleted when the job ends, the result can be downloaded for 60 minutes, and the stored copy is erased by a daily storage rule some time after it is 1 day old.
Step by step
- 1Open the PDF to Word tool and drop your scanned PDF where it says Add your PDF, or click to pick one. Files up to 100 MB are accepted.
- 2Tick “Recognise text in scans (OCR)”.
- 3Under Languages, select every language the pages are written in, up to 4. English is selected when the page opens, with languages your browser is set to where the picker offers them; deselect any the document is not written in.
- 4Press Export DOCX, which uploads your PDF over an encrypted connection to our server, where the OCR and the conversion run in one job.
- 5Download the DOCX when it is ready. The download link stops working after 60 minutes, so save it right away.
- 6Open it in Word, LibreOffice or Google Docs and read the text through against the page before you rely on it.
What the Word file holds
Each scanned page comes through as its picture, followed by the text read from it, as plain paragraphs you can edit, search and copy; a table read from a scan may come through as a Word table or as lines of text. Keep the picture while you check the text against it, and delete it when you no longer need it.
The look of the page is not rebuilt: fonts, columns and spacing are not carried over, and a line of the scan can arrive as a paragraph of its own, so expect to join some lines and style the headings yourself.
A PDF that mixes typed pages and scanned ones can go through with the switch on: pages that already have text are kept as they are, and only the scanned ones are read.
Choosing the languages
The language you pick decides the result. A page read in the wrong language does not fail: it comes back as letters that look like words and mean nothing, with no error, so select the language the scan is written in, and add a second one if the pages mix two.
The picker offers 30 of the 34 languages OCR PDF reads. Arabic, Hebrew, Persian, and Urdu are not offered here: written right to left, their recognised words come out of the conversion out of order. For those, use OCR PDF, which reads them in order in its searchable PDF and its text.
Getting a good result
OCR reads printed and typed text: letters, forms, contracts, book pages and office documents. It is not made for handwriting, which usually comes back as nonsense or not at all.
The clearer the scan, the fewer mistakes. Scan in your scanner's document or text setting rather than a low-resolution preview, keep the page straight and flat, and avoid shadows, glare and blur when you photograph a page with a phone.
A searchable PDF, or a photo instead of a PDF
If you need to find and copy the words but not edit them, OCR PDF keeps the page exactly as it looks and lays the recognised text invisibly under it.
If your scan is a JPG or PNG photo rather than a PDF, turn it into a PDF first with JPG to PDF, which runs in your browser, then convert that PDF here with “Recognise text in scans (OCR)” ticked.
Common questions
Can I convert a scanned PDF to Word for free?
Yes. There is no account and no payment: open PDF to Word, tick “Recognise text in scans (OCR)”, pick the languages, and download the DOCX.
Is OCR PDF a separate step?
No. “Recognise text in scans (OCR)” runs the OCR inside the same job, so one run of PDF to Word gives you the Word file. OCR PDF is for when you want a searchable PDF instead.
Why does my text come out as gibberish?
Because the page was read in the wrong language. English is selected when the page opens, with languages your browser is set to where the picker offers them; select the language your scan is written in, deselect any it is not, and run the job again. A blurry or sideways scan can cause the same symptom.
Can it read handwriting?
No. The OCR is made for printed and typed text; handwriting usually comes back as nonsense or not at all.
Is my file uploaded?
Yes. OCR needs a recognition engine that runs on our secure server, so your PDF is uploaded over an encrypted connection. The uploaded file is deleted when the job ends, and the result can be downloaded for 60 minutes.