Skip to content
EN

Extract text from an image (OCR)

Copy the text from a photo, screenshot or scanned PDF: recognition in 6 languages, editable text, export to .txt or Word. Runs in your browser.

  • Free
  • No sign-up
  • No file uploaded
  • Windows, Mac, Android, iPhone

Your files stay on your device: nothing is uploaded.

Loading the tool…

Share this tool

About this tool

A photo of a document, a screenshot or a scanned PDF holds text you cannot select. Drop the file: optical character recognition (OCR) reads it and gives you editable text to copy or save as .txt or Word. Everything happens in your browser: your document is never uploaded.

When to use it

Get the text of a paper document

A photographed letter, contract or invoice: get its text back without retyping it.

Copy text from a screenshot

A message, web page or caption you captured: copy it to quote, translate or search it.

Make a scanned PDF reusable

A PDF made of images cannot be copied: OCR turns it into text you can edit in Word.

How to use

  1. Select the document's language (several if needed).
  2. Drop an image (JPG, PNG, WebP…) or a scanned PDF.
  3. Wait while it is read, page by page.
  4. Fix the text if needed, then copy it or save it as .txt or Word.

How it works

Reading uses Tesseract, the open-source recognition engine, in its browser version Tesseract.js (Apache 2.0 license), with the “tessdata_fast” language data (Apache 2.0 license).

The engine and the selected languages are downloaded from zanakit.com on first use, then kept in the browser cache. PDF pages are rendered at 300 dots per inch before being read.

Your image or PDF is never sent to a server.

Tips for best results

  • Photograph the page flat, straight on and in good light, without shadows.
  • Text at least 10 pixels tall is recognized much better: move closer rather than zooming.
  • Only select the languages present: each extra language slows reading and can add errors.
  • To photograph a document and straighten it, start with “Scan to PDF with camera”.

Common problems and fixes

The text has many mistakes

Take a sharper, better-lit photo and check the selected language. Handwriting is barely recognized.

Columns or tables are mixed up

OCR returns text in the reading order it guesses. Crop the image to one column at a time with “Crop an image”.

A long PDF takes a while

Pages are read one after another. Extract only the pages you need with “Extract PDF pages”.

Frequently asked questions

Is my document uploaded to a server?

No. Recognition runs in your browser; only the engine and language data are downloaded from our site.

Which languages are recognized?

French, English, Spanish, Portuguese, German and Indonesian. You can select several for a mixed document.

How much data is downloaded?

About 5 to 8 MB the first time, depending on the language; after that, nothing: the engine and languages stay in your browser.

Is handwriting recognized?

Very poorly. Tesseract is made for printed or on-screen text.

Can I read a scanned PDF?

Yes, up to 50 pages at a time. If the PDF already contains text, “PDF to Word” keeps the layout better.

Is it free, even for commercial use?

Yes. Tesseract.js and the language data are under the Apache 2.0 license, and the text you get is yours.

Related searches

Still stuck?

Ask on the forum: the category, the tool, your device sheet and the test result will be filled in for you.