Image to text: extract the text from photos and screenshots
Choose one or more images and the language of the text: it is recognised on your device and you can correct, copy or download it. Your images never leave the browser.
Select images
or drag your photos or screenshots here
Preview of the first image as it will be recognised.
How to use it
OCR (optical character recognition) reads the letters in a photo or a screenshot and turns them into text you can copy, edit and save. Recognition runs with Tesseract inside your browser.
Choose one or more images with text: photos, screenshots or photographed documents.
Set the language of the text and, for a photo with poor lighting, turn on contrast improvement.
Press Recognise text, review the result and copy it or download it as .txt.
What is extracting text from an image (OCR) useful for?
Copying text from screenshots and whiteboard photos
Turn images into editable text without having to retype everything by hand.
Digitizing receipts, bills, and printed tickets
Extract figures and line items to paste into spreadsheets or accounting tools.
Recognizing text in multiple languages with local AI
Built-in Tesseract OCR engine processing images on your device without network uploads.
Copying text from books, signs, and lecture notes
Transfer photographed notes into Word documents or emails in seconds.
Why use JapalTools?
Your files never leave your device
Everything happens in your browser. Documents are never uploaded to any server, so you can use it with payslips, contracts or ID documents.
Free, no sign-up and no watermarks
You don't need an account or to leave your email, and nothing is added to the result.
Works on any device
Computer, tablet or phone, with an up-to-date browser (Chrome, Edge, Firefox or Safari). Nothing to install.
Frequently asked questions
Screenshots and sharp, well-lit, straight photos of printed text. With handwriting, blurry photos or the wrong language selected, the result is worse. Always review the text before using it, especially figures and names.
Yes. They are recognised one after another and the text is gathered into a single result, with each file name as a separator. The first time, the engine and the language data (several MB) are loaded from this same website, so it may take a moment.
No. Recognition runs in your own browser with Tesseract, and the engine and languages are served by JapalTools, not by third parties. Neither the images nor the recognised text are sent anywhere.
No. The document is processed in your own browser and is not sent to JapalTools or anyone else. When you close the page, no copy remains outside your device.