How to change text in a scanned PDF
Product preview of the Windows app. It cannot be downloaded or bought yet; the release date is to be announced. This page describes what the planned version is to do and where it applies.
Short answer: A scanned page is an image, so there is no text to edit until the words are recognised (OCR). Desktop editors such as Adobe Acrobat Pro and Foxit PDF Editor recognise the page and then let you retype words; the result depends on the scan quality and on having a matching font. Free browser tools can recognise and copy the text, but they cannot change the words on the scanned page itself.
The short answer
A scan has no text to edit: each page is an image. To change a word, a program must first recognise the page (OCR) and then rewrite the word so that it fits the picture around it. Paid desktop editors do both. Browser tools, including the free ones on this site, can recognise and copy the text, or add a searchable layer, but they do not change the words you see on the scan.
An example
Download the scanned registration letter (no text layer) and the expected result, in which “31 December 2025” has become “31 December 2026” and the page otherwise looks the same. If this is your situation — a date, a name or an amount on a scanned page needs to change — read on.
First, check what your file is
Open your PDF in the text-layer check. It shows for each page whether there is text:
- No text found (page has images) — often a scan, but not proof of one: a page that has an image and whose letters are drawn as shapes, or whose text cannot be decoded, gets the same result (without an image such a page shows No text found (drawings)). A broken character map can also give Text found while copying produces wrong characters, so copy a sample and compare it with the page. Look at the page and try to select a word in the main text; if nothing can be selected and the page is a picture of a document, continue with the methods below.
- Text found — the words are stored as text; you can usually edit them directly in a PDF editor. On Text and images pages, select a word in the main text first: the text found may be only a page number or a stamp on a scan.
- Image and text that is not displayed — a scan that already has an OCR layer. Editing the image is still needed, and the old OCR text must be updated too.
Check your file here
Choose the PDF: each page is checked for a text layer in this browser tab, without upload. Up to 100 MB and 1000 pages per file.
This tool needs JavaScript. If it does not start, enable JavaScript in your browser and reload this page.
This browser cannot run the tool. Please use a current version of Chrome, Edge, Firefox or Safari.
No file selected yet.
What works today
Adobe Acrobat Pro (paid, Windows and macOS)
Steps from Adobe’s help page Edit scanned documents (last updated 22 October 2026):
- Open the scanned PDF.
- Select Edit in the top menu. Acrobat applies OCR automatically and creates an editable copy.
- Select the text element and type. Adobe states that the new text matches the original font of the scan.
- Save with Menu > Save as (Windows) or File > Save as (macOS).
Adobe’s own tips: check the recognised text before editing and run Scan & OCR > Recognize Text again if it is poor; improve blurry scans first; and, from Change, replace, or delete text: to edit with a font, that font must be installed on your computer — if it is only embedded in the PDF, you can change only the colour or size of the text.
Foxit PDF Editor (paid, Windows and macOS)
From Foxit’s product pages: run Convert > Recognize Text, choose the languages and pages and set the output to Editable Text (the other output, Searchable Image Text, only adds searchable text behind the unchanged scan, so the visible words cannot be edited), then use Edit > Edit Text and type. Foxit highlights words it is unsure of after recognition; check them before editing.
Free workarounds, with their cost
- Recognise, then rebuild the document. Get the text with PDF OCR, correct it in a word processor and make a new PDF. The text is right, but the page is no longer the original scan.
- Cover and retype. In an image editor, paint over the old word on the page image and type the new one, then put the image back into a PDF (JPG/PNG to PDF). Done this way the old pixels are gone; drawing a box in a PDF editor instead leaves them underneath.
What the browser tools on this site can do here
They help with the steps around the edit, not the edit itself:
- Text-layer check — is it a scan at all?
- PDF OCR — get the text out to copy, check or rebuild it.
- Make PDF searchable — after editing, give the pages a correct text layer again.
- Compare PDF text — compares the text layers of two versions. Both files need one, so make an image-only original searchable first. Recognition errors show up as differences too, and the scan image itself is not compared: look at every edited page as well.
The Windows app (product preview)
This is a preview of a Windows app: it cannot be downloaded or bought yet, and the release date is to be announced. The planned version is to replace words on scanned pages in place: it recognises the page on your computer, erases the old strokes in the page image, writes the new text in the same position and size, and updates the text layer, then reads the saved file back to check the change. It is planned for scans with a clear, straight print; handwriting, heavily degraded scans and right-to-left scripts are outside its scope.
How to check the result
- Look at the changed page at 200 % or more: the new word should sit on the same baseline, in a similar weight, with no leftover strokes around it.
- Search the file for the old word (Ctrl+F / Cmd+F) and look at the matches on the edited page. A match at the place you changed means its text layer was not updated; matches elsewhere can be occurrences you did not change.
- Search for the new word. If it is not found, check the search options and try to select or copy the word first: a search can also miss it because of word breaks or character mapping. Only when the page clearly has no matching text layer, use Make PDF searchable with “Redo page”: it turns the page into an image and recognizes its text again, replacing the old text layer.
- Compare the old and new file with Compare PDF text as an extra check of the text layer: it lists changed words, but ignores differences in spacing and treats superscripts and full-width letters as their plain forms, so steps 1–3 remain necessary.
FAQ
Why can't I click into the text of my scanned PDF?
The page is a photo of the paper. A PDF viewer can only edit text that is stored as text. Until the page is recognised (OCR), there are no words in the file, only pixels.
Is drawing a white box and typing new text on top enough?
It changes what people see, but the old words usually stay: in the page image under the box, and in any OCR text layer, so searching or copying still finds them. Use it only for documents where that does not matter.
Will the edited words look like the rest of the page?
Only if the editor has a similar font and the scan is clean. Acrobat states that the new text matches the original font; it also notes that a font must be installed on your computer to edit with it. Check the result at full size.
Sources
- Adobe: Edit scanned documents (Acrobat desktop help, last updated 22 Oct 2026) · Checked on 2026-11-12
- Adobe: Change, replace, or delete text (Acrobat desktop help, last updated 31 Aug 2026) · Checked on 2026-11-12
- Foxit PDF Editor: Scan to PDF and OCR · Checked on 2026-11-12
- Foxit PDF Editor: Advanced editing · Checked on 2026-11-12