How to turn a scanned PDF into an editable Word file without retyping it

You have a scanned contract, a photographed letter or a faxed report, and you need the text in Word: to reuse a clause, to update a template, to quote three paragraphs in an email. Retyping is the last resort. This guide shows how to get an editable Word file out of a scan, what to check before you trust it, and why the PDF you get on the way is often the better deliverable.

Why a scanned PDF has no text to copy

A scanner produces an image. When that image is saved as a PDF, the file contains one picture per page and nothing else: no characters, no words, no fonts. Select-all selects nothing, search finds nothing, and "Save as Word" in a viewer produces a document with one big picture on each page. Before a scan can be edited or converted, the pixels have to be turned back into text. That is optical character recognition, and the quality of the result is decided there.

What good recognition returns

Modern document recognition does more than read letters. It returns each word with its position on the page, the measured type size, whether it is bold, italic or handwritten, and its colour, and it finds the checkboxes. blupdf uses that to rebuild the page as a real document rather than laying invisible text over the picture: every line is drawn again as real glyphs in the closest matching font at the measured size, rules become vector lines, checkboxes become boxes with their marks, and logos and photos are cropped from the scan and put back where they were. The page you see is a document, and it is what both downloads are made from.

Step by step

1. 1

   Open the scan in Studio

   Drop the PDF, or the photo itself (PNG, JPG and WebP are wrapped into a page). blupdf checks the file against your plan's page and size limits and opens it.

2. 2

   Let each page be recognised

   A scanned page is recognised the first time you reach it; Reading the text on this page takes a few seconds, then Rebuilding this page draws the real-text version. The charge is 80 credits per page, once.

3. 3

   Check the doubtful words

   Click Show original scan on a page to compare the rebuilt text with the picture. The classic misreads are 0 and O, 1 and l and I, and rn read as m. Fix any of them by clicking the block and retyping; it is an ordinary edit.

4. 4

   Download as DOCX

   Choose Download, switch the format to DOCX and name the file. Word opens it as paragraphs of editable text in reading order, with the size, weight, slant, colour and alignment of each block.

5. 5

   Or download as PDF

   Keep PDF and you get the rebuilt document: real, selectable, searchable text in matched fonts, with the layout intact. For most recipients this is the better file.

Word copy versus PDF: which to send

The DOCX download is a text copy, not a layout clone. It gives you every text block as a paragraph, top to bottom and left to right, with its styling, and it leaves out pictures and signatures. That is what you want when the text is going into another document. When the layout matters, when the file is going to someone who will read it rather than reuse it, or when it needs to be searchable in an archive, download the PDF: it keeps every line where it was, in a matched font, as real text. Downloads are free in both formats, so take both.

The scan-quality rules that decide accuracy

Recognition is very accurate on clean office scans and noticeably less so on poor input. Four things matter most:

• Resolution. 300 dpi is ideal. Below 150 dpi, small text and punctuation blur together, and 6-point footnotes are the first casualties.
• Skew and perspective. A page photographed at an angle stretches letters unevenly. Scan flat; if you must photograph, hold the phone parallel to the paper.
• Lighting. A shadow across the page loses characters in the dark band. Photograph near a window, not under a single lamp.
• Handwriting. Printed text is reliable. Handwritten entries are recognised and flagged as handwriting; read those lines against the original before relying on them.

Language is rarely the problem: printed text in most languages and scripts is recognised, and mixed-language pages are fine.

Editing before you convert

Because the rebuilt page is a real document, you can change it before downloading. Retype a wrong date in the matched font, delete a paragraph and the paper tone stays clean, or open blupdf AI and say what should change: "change the client name to Acme GmbH everywhere" lands in every place it occurs, in the same font, for 100 credits per request. The DOCX and PDF you download include those edits.

Cost

Recognising a scanned page costs 80 credits, once per page, charged after the page is read. Text you retype costs 1 credit per character; downloads cost nothing. A ten-page scanned report converted to Word with a dozen corrections costs about 800 + 50 = 850 credits. Every account starts with 1,000 credits, no card required.

FAQ

Does the Word file keep the layout?

No. It is a faithful copy of the text with each block's styling, in reading order. For the layout, download the PDF, which keeps every line where it was as real text.

Can I convert a photo of a page, not just a PDF?

Yes. Drop the PNG, JPG or WebP into Studio; it is wrapped into a page sized to the picture and recognised like a scan. Crop to the page first for the best result.

Is the original scan kept?

Yes. Show original scan on any page switches between the rebuilt text and the scan, and the original file stays in your account.

The reference is Editing scanned PDFs with OCR; the download formats are explained in Pages, downloads and your library.