DigiTools

Nothing uploaded

Extract text from an image

Turn photos, scans and screenshots into text you can copy, edit and search. Reads English, Spanish, Portuguese and other languages written in the Latin alphabet. The AI runs in your browser, so your images stay on your device.

or drop them anywhere on this page

Files stay on this device. Nothing is uploaded.

How to extract text from an image

  1. Choose your images, or drop them onto the page.
  2. Select Extract text.
  3. Copy the text, or download it as a .txt file (a ZIP for several images).

Questions

Is it really free, and are my images uploaded?

It’s free, with no sign-up. And nothing is uploaded: the AI runs inside your browser on this device, so your images and their text are never sent to a server.

Which languages can it read?

Languages written in the Latin alphabet, accents included: English, Spanish, Portuguese, French, German, Italian, Dutch and many more. It doesn’t read other alphabets such as Chinese, Arabic or Cyrillic yet.

How accurate is it?

Very accurate for clear printed text: screenshots, scans and photos taken straight on. Small, blurry, curved or decorative text, and handwriting, give more mistakes, and lines it wasn’t sure of are marked with a dashed outline. Check names and numbers that matter.

Can it read a PDF?

Not directly yet. Turn the PDF’s pages into images with PDF to JPG, then extract the text from them here. A PDF that already has text in it doesn’t need this: you can copy its text in any PDF reader. A tool that makes scanned PDFs searchable may follow.

Why does it download something the first time?

The first time you use it, your browser downloads the AI models, about 13 MB, from this site, and keeps them for next time. The models only do the work: your images stay on your device.

Which AI models does it use?

PP-OCRv5 from PaddleOCR by PaddlePaddle (Baidu): its mobile text detection model and its recognition model for the Latin alphabet, released under the Apache 2.0 licence.

Need to do more?