How to convert a scanned document into editable Word text

Copy a paragraph from a scanned PDF and you get nothing, because there is nothing to copy: the page is a photograph. Optical character recognition, or OCR, looks at the shapes of the pixels and works out which letters they are. Done well it turns a filing cabinet into searchable, editable text.

Step 1 — start with the best possible scan

• Scan at 300 dpi. Below about 200 dpi, small print becomes guesswork for the recogniser.
• Keep the page straight. A few degrees of skew noticeably lowers accuracy; most scanner apps have a deskew or auto-crop option.
• Prefer greyscale or black and white over colour for plain documents — it removes distracting background tint.
• Photograph in even light with no shadow across the page if you are using a phone.

Step 2 — run the recognition

1. Open the PDF OCR to Word tool and choose the language your document is written in — this matters, because the recogniser uses a dictionary as well as letter shapes.
2. Add your scanned PDF. Each page is rendered at high resolution and read in turn.
3. Watch the progress bar. The first run downloads the recognition engine and language data once, so it takes a little longer than later runs.
4. Read the recognised text on screen, then download the .docx.

Everything happens inside your browser tab. Scanned passports, payslips and contracts are never uploaded to a server.

Step 3 — proofread properly

Never treat OCR output as finished. Check numbers first — digits are where errors hurt most, and a mistaken 8 for a 3 in an invoice is expensive. Then search the document for the letters most often confused: l and 1, O and 0, rn read as m. A five-minute pass catches nearly everything.

What OCR will not do

• Reproduce your layout. Columns, tables and images are not rebuilt — you get the text.
• Read handwriting reliably. Cursive and hurried notes are beyond this kind of recognition.
• Rescue a very poor scan. If you cannot read a word yourself on screen, neither can the recogniser.

If your PDF was created digitally rather than scanned, skip OCR entirely — PDF to Word extracts the existing text layer directly and gives a cleaner, faster and perfectly accurate result.