PDF to Word ConverterHow to Convert a Scanned PDF to Editable Word — Free
A scanned PDF is a photograph of text, so no ordinary converter — ours included — can turn it into editable Word. You need OCR, and the genuinely free ways to get it are Google Drive's built-in recognition, the open-source Tesseract engine, or your iPhone's Live Text.
Why scanned PDFs won't convert to Word
A scanned PDF is a picture of a page. There are no letters inside the file, just pixels arranged to look like letters. A normal PDF-to-Word converter works by reading the PDF's text layer, and a raw scan has none, so the converter has nothing to extract. That includes the converter on this site: it reads the text layer only and does not do OCR.
Run a scan through most free converters and you get a blank document, or a Word file with the page embedded as one big image you still can't edit. Turning pixels back into real characters requires OCR (optical character recognition), which is a much harder problem. That's why most free tools skip it and most big services charge for it.
First, check whether your PDF is actually scanned
Plenty of PDFs that came from a scanner still have a text layer, because the scanning software ran OCR before saving. The quick test: open the PDF and try to select a sentence with your cursor. If you can highlight and copy the words, there's a text layer and an ordinary converter will work. If your cursor just draws a box over the whole page as if it were a photo, it's a raw scan.
The converter below runs this check for you. Drop the file in — if there's no text layer, it tells you so instead of handing you an empty document. If there is one, it rebuilds paragraphs, headings, bold and italic into a proper DOCX. The same detection runs in our PDF to text extractor.
OCR routes that are genuinely free
The best-known free route is Google Drive. Upload the scanned PDF, right-click it, choose Open with → Google Docs, and Drive runs OCR and opens the recognized text as an editable document. Download it as Word via File → Download → Microsoft Word (.docx). Recognition is solid on clean, straight scans; formatting mostly isn't preserved, so expect plain paragraphs.
The trade-off: your file goes to Google's servers, which may rule this out for contracts or medical records — see our guide on online converter safety. Technical users can run Tesseract, a free open-source OCR engine, entirely on their own machine.
The big converter sites do offer OCR, but usually behind a paid tier — Smallpdf lists it under its Pro plan and iLovePDF includes it in Premium, per their own pricing pages. That's not a knock; OCR costs real compute. It just means the free buttons on those sites won't recognize a scan.
Make a scanned PDF editable on iPhone
If the document is still on paper, skip the scanned-PDF problem entirely. Open Notes, tap the camera icon, and choose Scan Documents. On recent iPhones, iOS recognizes the text in the scan, so you can press and hold to select it, copy, and paste into Pages, Word, or an email.
For a scanned PDF you already have, a workable per-page trick: screenshot each page, open the screenshot in Photos, then press and hold the text and copy it with Live Text. Tedious for long files, but free and fully on-device.
Once a PDF does have a text layer, our free iPhone app converts it to a real .docx and saves it straight to Files. It runs on iPhone and iPad (iOS 15.1+), keeps a conversion history, and costs nothing.
What to expect from any OCR result
Even good OCR isn't perfect. Expect occasional character mix-ups (l vs 1, O vs 0), especially on low-resolution or skewed scans. Tables often collapse into plain lines, and multi-column layouts can interleave. Proofread anything that matters — numbers in contracts and invoices especially.
If your file already has OCR applied, it's a normal PDF with a text layer, and the converter on this page will turn it into an editable Word file whose text reflows properly when you edit. To keep fonts and layout as intact as possible during that step, see converting PDF to Word without losing formatting.