How to Copy Text from a Scanned PDF

August 2, 2026

You cannot copy text from a scanned PDF because there is no text in it — only a picture of text. The fix is OCR: software reads the image and reconstructs the words, which you can then select, search, and paste. It takes one pass in your browser, and the document never has to be uploaded anywhere.

How to copy text from a scanned PDF

  1. Open the OCR PDF tool and drop in your scanned document.
  2. Choose the document language — English, Indonesian, or both together for a mixed-language file.
  3. Tap Recognize Text. Pages are processed one at a time on your device, with the page number and percentage shown as it goes.
  4. The searchable PDF downloads automatically. Open it and you can now select and copy text directly, exactly as in a normal document.
  5. Prefer the raw words? Use Download extracted text (.txt) to get everything as plain text, ready to paste into an editor.

The two outputs suit different jobs. The searchable PDF keeps the page looking exactly as scanned — signatures, stamps, letterhead all intact — with an invisible text layer sitting on top. The .txt file throws the layout away and hands you the content, which is what you want when the goal is to quote a paragraph or move the contents into a spreadsheet.

Why the text was not selectable in the first place

A PDF can hold two very different kinds of page. A document exported from a word processor stores its text as characters plus font instructions, so selecting and searching just work. A scan — from a flatbed, an office copier, or a phone camera — stores one flat image per page. Your eyes read words; the file contains pixels. That is why the cursor refuses to select anything, why Ctrl+F finds nothing, and why the file is so much larger than a text document of the same length.

OCR (optical character recognition) closes that gap by analysing the image and working out which shapes are which characters. It does not alter the scan itself — see our companion guide on making a scanned PDF searchable for what the added text layer looks like under the hood.

Getting better results out of OCR

Recognition quality depends far more on the scan than on the software. A few things make a real difference:

  • Resolution. Around 300 DPI is the sweet spot for printed text. Much lower and characters blur together; much higher mostly adds processing time.
  • Straightness.Skewed and curled phone photos cost accuracy. Use your phone’s document scanner, which flattens and de-skews the page, rather than a plain camera shot.
  • Contrast. Crisp black on white beats a grey photocopy of a photocopy. Faint or speckled scans are where errors cluster.
  • The right language. An Indonesian document processed as English will mangle exactly the words that carry the meaning.

Always read the output before relying on it. OCR is very good, not infallible — the characters worth double-checking are the ones that look alike in print: 0 and O, 1 and l, rn and m.

Nothing about your document is uploaded

Scanned documents are almost by definition the private ones: contracts, ID cards, medical letters, bank statements. Most online OCR services upload the file to a server, run recognition there, and send the text back — so the whole document, and the text extracted from it, passed through someone else’s system. PDFMatic runs the recognition engine inside your browser via WebAssembly, so the scan stays on your device. Verify it for yourself: load the page once, disconnect from the internet, and run OCR anyway — it works, because there is no server to reach.

Copying text from a scanned PDF — FAQ

Can I paste the recognised text straight into Word or Google Docs?

Yes. Alongside the searchable PDF you get a plain .txt file of everything that was recognised, which pastes cleanly into any editor. Expect to fix the layout afterwards: OCR recovers the words, not the original columns, tables, or styling.

Does it work on handwriting?

Not reliably. The recognition engine is trained on printed type, so neat block capitals sometimes come through while ordinary cursive mostly does not. Printed and typewritten documents are where OCR earns its keep.

Why is the first run slower than the rest?

The first time you use it, your browser downloads the recognition engine — around 6 MB — and caches it. Every run after that starts immediately, including offline. Recognition speed after that depends on your device and the number of pages.

My document is a photo, not a PDF. Can I still use it?

The tool reads PDFs, so wrap the photo in one first. On iPhone and Android the Files or Photos app can save an image as a PDF in a couple of taps, and phone scanner features produce a PDF directly — that file is then ready for OCR.

Which languages can it recognise?

English and Indonesian, and you can enable both at once for documents that mix the two. Picking the wrong language is one of the most common causes of poor results, so it is worth checking before you run it.

Try it on your own scan: OCR PDF is free with no page limits, watermarks, or account. If the resulting file is bigger than you would like, the Compress PDF tool will trim it back down.