Copy text from scans (OCR)

How to copy text from a scanned document on Android, offline

Short answer: To copy text from a scanned document on Android, open the scan in Scanly and use its Extracted text screen. Scanly reads the words on the page with on-device OCR, so nothing is uploaded and you can copy, share or edit the text.

A scanned page or a photo of a document is an image, so you cannot select or copy its words the way you would with a digital file. Optical character recognition (OCR) reads the letters in the image and turns them into text you can use.

Scanly uses ML Kit Text Recognition v2 with a model bundled in the app. Recognition runs on your phone, so the text of your receipts, letters and contracts stays on the device.

Install Scanly free on Google Play (Android 10+, no account, no ads)

How to copy text from a scanned document on Android

  1. Open the document

    Open the scanned document from your Scanly library. To work on a photo or PDF from elsewhere, import it first; Scanly can import up to 50 images at a time.

  2. Open the Extracted text screen

    Open the Extracted text screen for that document.

  3. Pick a view

    Use Text view to read the recognised text as a block, or On-page view to check it against the page layout.

  4. Copy the text

    Copy the text you need, or copy all pages in one go.

  5. Share or edit

    Share the text through the Android share sheet, or edit it first to fix mistakes before you reuse it.

  6. Check names, numbers and dates

    Compare the copied text with the page, especially for amounts, dates and names. OCR is useful but not perfect, and a quick check prevents errors.

What you can do with the recognised text

  • Search inside your scans. Find a document by words printed on its pages, not only by its file name.
  • Copy text from a page, or copy all pages at once.
  • Use Text view to read the recognised text as one block, or On-page view to see how the text sits on the page.
  • Share or edit the extracted text before you reuse it somewhere else.

Which documents work best

Recognition works best on printed text with good contrast and a sharp scan. Scanly is built around common document types, including invoices, receipts, contracts, statements, letters, certificates, ID documents, forms and résumés. Handwriting, unusual fonts and blurry photos are harder, so compare the result with the page before you rely on it.

What this does not do

Text recognition in Scanly is for searching your library and copying text. Exported PDFs are page images, so do not rely on selecting text inside them. If you need the text from many pages, use Copy all pages on the Extracted text screen and paste it into your own document.

Questions people ask

How do I copy text from a scanned document on Android?
Open the document in Scanly, go to its Extracted text screen, and copy the text or copy all pages. Recognition runs on your phone and nothing is uploaded.
Does OCR work offline on Android?
Yes. Scanly's text recognition uses a model bundled in the app, so it works without an internet connection.
Does Scanly upload my scanned text to a server?
No. Text recognition is on-device and nothing is uploaded for it. Files leave your phone only when you export or share them.
Can I search my scans by the text on the pages?
Yes. Scanly lets you search your library by document name and by words printed on the pages.
Can Scanly make a searchable PDF?
Not today. Scanly uses on-device text recognition to make your library searchable and to let you copy or share the text from any page, but exported PDFs do not include a selectable text layer.

Last checked 2026-10-22. Markdown version