Convert a scanned PDF to editable Word

Upload a scan or a photo of a page. Text recognition (OCR) turns it into Word text you can edit, search and copy.

  1. Upload
  2. Choose pages
  3. Convert

Upload a PDF

or drop a PDF or image here

No account or card required · PDF, JPG or PNG up to 50 MB · scans welcome · not password-protected

Advanced options Optional

Your free page uses the standard settings. On the result screen you can still turn on Keep page breaks before downloading. With a free account you can also choose it before converting.

Files from your free page are deleted 24 hours after you convert it. Scanned and complex pages are read by Microsoft.

Page 1

From a picture of text to text you can edit

A real scan, converted and unedited

A typed 1969 legal notice from a public archive, with a stamp, a signature and a numbered list. This is exactly what came back.

The scanned notice: a typed transmittal form with address blocks, a round Received stamp, a numbered list of nine items about a lawsuit, a small grid of initials and a signature.
Before: the scan, an image with no selectable text
The first page of the converted notice in Microsoft Word: the address blocks as text, the numbered items 1 to 7 as a Word list, and the grid of initials as a table.
After: the .docx opened in Microsoft Word (page 1 of 2)
What came through, and what needed fixing

Came through as editable Word

  • The typed text as ordinary paragraphs, such as “Reuben Sunstein, Plaintiff, v. The American Tobacco Company, a New Jersey corporation, Defendant.”
  • Items 1 to 9 as one real Word numbered list
  • The grid of initials as a Word table
  • No text boxes or floating frames, in one Word section

Needed a quick fix by hand

  • Some words misread: “coples” for “copies”, “Courl” for “Court”, “Servient” for “Service:”, and “$5, 000” with a space
  • Small printed labels under the lines read badly, such as “(Stole)” for “(State)”, and the letterhead’s “C T SYSTEM” came out as “BT SYSTEM”
  • The stamp and the signature became stray text (“VAX DEPT.”, “V. Colorce”)
  • The address block at the top is spread out and partly out of order, and the page runs onto a second Word page

Check it yourself: ·

Questions about scanned PDFs

Do I need to run OCR on the PDF first?

No. Upload the scanned PDF as it is: the text is recognised during the conversion. Scanned pages are read by Microsoft; see .

Can I convert a photo of a document?

Yes. A JPG, PNG, WebP, GIF or BMP image is converted as one page, up to 50 MB and between 50 and 10,000 pixels on each side. TIFF and HEIC aren’t supported yet: save the image as JPG, PNG or PDF first.

How accurate is the text?

It depends on the scan. Clean, typed pages come through best; faint, smudged or low-resolution pages and handwriting need more checking, as the example above shows. Flag scanning errors (on by default) lists words that may be misread, with a suggestion when there is one.

Will the Word file look like the scan?

Not exactly. You get a clean document that flows like a normal Word file, and fonts and spacing can change. To start each scanned page on a new Word page, turn on Keep page breaks in the conversion settings.

Convert your own scan

Your first page is free. No account, no card.

Choose a file