Getting text out of scans

Searchable PDF, how to check yours and add a text layer free

What makes a PDF searchable, a five-second test to see if yours is, free ways to add a text layer, and why copied text from a scan comes out wrong.

Updated 20 November 2026Checked against Google Drive Help, NAPS2 8.6.0, OCRmyPDF 17.14.1, iOS 27, Adobe online OCR, Nov 2026

Short answer

A searchable PDF holds real text that Find and copy can read, and on a scan that text is a hidden OCR layer laid over the page image. To check yours, press Ctrl+F (Cmd+F on a Mac) and search for a word you can see. To add the layer free, run the file through NAPS2 on a computer.

The catch is that a match in Find only proves the layer exists. OCR software guesses the text from the image, and search and copy-paste use that guess, whatever the page shows.

What makes a PDF searchable

A PDF saved from a document on your computer, such as from Word, is typically searchable as standard, because the text was text from the start. A scanned or faxed page arrives as a picture, so it isn't searchable until something reads it. That something is optical character recognition (OCR). It reads the image, works out the words and their positions, and writes them into the PDF as text you can't see, so the page looks unchanged.

Test whether your PDF is searchable

  1. Open the PDF on a computer, in your browser or any PDF viewer.
  2. Press Ctrl+F on Windows or Cmd+F on a Mac, and type a word you can see on the first page. Use a word that's clearly printed, so a missing match points to no layer rather than one misread word.
  3. Look at the result. If the word is found and highlighted on the page, the PDF has a text layer. If there are no results, you have an image-only scan, which Adobe describes as merely an image of the copied document.
  4. Press Ctrl+A (Cmd+A on a Mac), copy, and paste into a plain text note. This shows what the layer actually says, and it's the same method NAPS2 suggests for getting OCR text out.

If the pasted text is clean, nothing needs doing. If it's garbled, see the last two sections.

Add a text layer free

Route Where it runs What you get Where the file goes Catch
NAPS2 Windows, Mac, Linux Searchable PDF Processed on your computer A download (a portable version runs without installing), plus a language download the first time
OCRmyPDF Command line on a computer Searchable PDF/A by default, which tends to make larger files Processed on your computer No buttons, only typed commands
Adobe Acrobat online OCR Web browser Searchable PDF Uploaded to Adobe's servers You have to sign in to download the result. The file is deleted unless you sign in to save it, and Adobe's page doesn't say how long a saved file is kept
Google Drive scanner Drive app on iPhone and Android Searchable PDF Your Google Drive Works on new paper scans only, not files you already have
Google Drive, Open with Google Docs Drive in a computer browser Editable Google Doc, not a PDF Your Google Drive The file should be 2 MB or smaller

Use NAPS2 for a scan you already have. It's free and open source, it reads over 100 languages, and the file doesn't need to go to anyone's servers. For payslips, bank statements and ID, that last point decides it. Adobe's tool suits a computer where you can't download or run programs at all. The Google Docs route only makes sense if you want text to edit rather than a searchable PDF, and Google warns the formatting might not transfer.

These steps follow NAPS2's OCR guide, with version 8.6.0 current on its site in November 2026.

  1. Download and install NAPS2 from naps2.com.
  2. Click OCR in the toolbar (on a Mac, Tools then OCR).
  3. The first time, NAPS2 asks you to pick languages. Tick the language your document is in and click Download, because OCR reads each language with its own data.
  4. In the OCR Setup window, check that Make PDFs searchable using OCR is ticked and the right language is selected. Every PDF you save from then on gets a text layer.
  5. Import your scanned PDF into NAPS2. With OCR enabled, NAPS2 adds OCR to each page that has no text.
  6. Click Save PDF, then pick a location and file name.
  7. Run the Find test above on the new file before you delete the old one.

If the scan is poor, NAPS2 has two settings in OCR Setup that can help. Best mode can give better results than Fast but is slower. Fix white balance and remove noise can improve OCR on low-quality scans.

Steps for

Make a searchable scan on iPhone

The steps for scanning with Notes or Files are in our guide to scanning documents on iPhone. Apple's iOS 27 guide to scanning in Notes says the scan is saved as a PDF but says nothing about a text layer, so run the Find test on one before you rely on it. Google says the Drive app's scanner saves searchable PDFs. In the Google Drive app, tap Add, then Scan. The PDF is saved to your Google Drive, so you need a Google account.

Make a searchable scan on Android

The full scanning steps are in our guide to scanning documents on Android. Google says the Drive app's scanner saves searchable PDFs. When you save, pick .pdf, because a .jpg is a picture and can't carry a text layer.

Why copy-paste from a searchable scan comes out wrong

Redo a bad text layer

Rerunning a garbled PDF through either tool changes nothing. NAPS2 leaves alone any page that already has text, bad OCR included. OCRmyPDF stops with "page already has text!" by default, and its docs say it can't quite tell printed text from a hidden OCR layer.

The error message suggests --force-ocr, but try --redo-ocr first. It removes the existing hidden OCR text and runs OCR again, leaves printed digital text alone, and replaces OCR without rasterising the pages. Add -l with the right language code, as in the cookbook's examples. Language packs must be installed for every language you name.

ocrmypdf -l eng --redo-ocr input.pdf output.pdf

Don't add image processing options such as --deskew, because --redo-ocr is incompatible with image processing options, which change how the page looks.

If that doesn't work, OCRmyPDF's --force-ocr turns every page into an image and runs OCR on it, which can reduce quality or lose vector content. Adobe doesn't say whether its online OCR replaces an existing text layer. Without the command line, the fix is to scan the paper again, more sharply, and add the text layer to the new file.

More on getting text out of scans

All getting text out of scans guides
How we made this guide

Carbonado Apps wrote this guide with the help of AI research tools. We checked it against Google Drive Help, NAPS2 8.6.0, OCRmyPDF 17.14.1, iOS 27, Adobe online OCR, Nov 2026, using the sources listed below. We make CrispCopy, and we wrote this to be useful whether or not you ever use it. If something here is wrong or out of date, email hello@carbonado.app and we'll correct it.

Sources (12)
  1. Searchable PDFAdobe
  2. OCR a PDF for freeAdobe
  3. Introduction (ocrmypdf 17.14.1 documentation)OCRmyPDF
  4. Common error messages (ocrmypdf 17.14.1 documentation)OCRmyPDF
  5. Cookbook (ocrmypdf 17.14.1 documentation)OCRmyPDF
  6. NAPS2, Scan documents to PDF and moreNAPS2
  7. NAPS2 OCRNAPS2
  8. Windows ScanningNAPS2
  9. Convert PDF and photo files to textGoogle
  10. Scan documents with Google Drive (Android phone and tablet)Google
  11. Scan documents with Google Drive (iPhone and iPad)Google
  12. Scan text and documents in Notes using the iPhone cameraApple
CrispCopyA document scanner and PDF maker for iPhone and Android with no catch.
Get it on Google Play