Make scanned PDFs searchable and selectable with OCR.
Open tool
Workflow page
OCR then Word
How scanned PDFs become editable Word files: recognize the text first, then convert. Includes alternatives if you have the original digital file.
The problem this solves
A scanned PDF is a stack of photographs of pages. Your eyes see words, but the file contains only pixels. If you send that file straight to a Word converter, there is no text for it to pull out, so you typically get a Word document with each page embedded as an image: nothing to select, search, or edit.
OCR (optical character recognition) reads those pixels and adds a real text layer. Only after that step does a conversion to Word have actual words to work with. That is why the order matters: recognize first, convert second. Reversing it gains you nothing.
How the two steps fit together
- Recognize the text. OCR PDF processes the file and returns the
result in the same request: you choose OCR languages (the installed ones
are English, German, French, Portuguese and Simplified Chinese:
eng,deu,fra,por,chi_sim; for exampleeng,deu), leave options such as Deskew scanned pages on, and download the searchable PDF. - Check the recognized text. Open the PDF and try selecting or searching a few words. Recognition errors are easier to spot here than after formatting is rebuilt in Word.
- Convert to Word. Upload the searchable PDF to PDF to Word and pick a Word format: DOCX (the default), DOC, or ODT.
- Proofread in Word. Names, figures, and dates deserve a line-by-line check against the scan, because OCR can misread similar characters.
If you may not need OCR
Before you run OCR, these routes can save time:
- Go back to the digital original. Many “scans” exist because someone printed a Word file and scanned it. Ask the sender for the original DOCX, or for a PDF exported directly from it. A PDF that was created digitally already contains text, and PDF to Word can convert it without any recognition step.
- Check whether the PDF already has text. Some scanners run OCR themselves. If you can select words in your PDF viewer, skip straight to PDF to Word.
- Use the scan as a reference, not a source. For a short document, retyping while looking at the scan can be quicker than correcting a poor conversion.
Note that Scan to PDF and JPG to PDF build image-based PDFs. They are good for capturing paper, but their output is not searchable text on its own.
An example
A small firm receives a signed services agreement as a scanned PDF and wants to reuse the clauses for a new client. The first thing to try is asking the other party for the editable original. If only the scan exists, the OCR step adds a text layer, the team checks a few clause numbers by searching for them, then converts to DOCX and proofreads against the scan before reusing any wording.
Troubleshooting
- Word file shows only images. The PDF has no text layer yet; it needs recognition, or a digital original.
- Strange characters or broken words. The scan may be skewed, low resolution, or in a language not set in OCR languages. A cleaner, straight scan helps more than any setting.
- Layout looks different from the scan. Conversions rebuild formatting approximately. Columns, tables, and stamps often need manual tidying.
When not to convert at all
If you only need to read, sign, or annotate the document, keep it as a PDF. Converting to Word is worth it when you need to change the wording. For more detail, see convert a scanned PDF to Word and PDF to Word for scanned contracts.
Trust center
Trust pages
How We Want PDF handles your files: privacy, security, and how long uploads are kept.
privacy
Privacy
How We Want PDF handles the files you upload, what is needed to use the tools, and which tools send content to an AI provider.
security
Security
How We Want PDF checks uploads, delivers results, limits abuse, and removes queued job files.
retention
File retention
How long We Want PDF keeps uploaded files and results, and when they are removed.