How to extract Arabic and English text from an image
The text you need is often trapped inside a picture: a screenshot of a message, a book page photographed with your phone, an image-based advert or a scanned invoice. Retyping it takes time and invites mistakes. This tool uses optical character recognition (OCR) to read Arabic and English text and turn it into plain text you can copy, paste and edit.
It runs the open-source Tesseract engine inside your browser. On first use the browser downloads the engine files and the Arabic and English language data from public distribution servers, but your image itself stays on your device and is not uploaded.
How to use it
- Drag an image onto the upload box or click it to choose a file (for example JPG or PNG).
- Click Extract text. A message explains that the engine is loading and then reading the image; the first run can take around a minute.
- The recognised text appears in the large read-only box below the button.
- Click Copy text to put it on your clipboard, then paste it into any document or message and edit it there.
What the tool offers
- Two languages at once: Arabic and English are recognised in the same image with no language setting to choose.
- One-click copy: the copy button appears only when text has been found.
- Clear feedback: if no clear text is detected, a message suggests trying a sharper image.
- No sign-up: repeat the process on as many images as you like, one after another.
Practical uses
- Copying a paragraph from a photographed book page for research notes.
- Pulling an order number or address out of a chat screenshot instead of typing it.
- Turning the text of an image-based post into editable text you can translate.
- Moving details from a scanned document into a Word file or spreadsheet.
Tips for better accuracy
- Clarity first: photograph pages in good light, straight on, without shadows across the text.
- Crop out the clutter: an image containing only the text gives cleaner results than one full of graphics and busy backgrounds.
- Printed text works best: clear printed fonts are read far more accurately than handwriting or decorative lettering.
- Always proofread: OCR can misread characters, especially Arabic dots and diacritics, so check the result before relying on it.
- Internet connection: you need to be online the first time so the engine can load; later runs on the same browser are usually faster.
If the text is inside a scanned PDF use Extract text from PDF (OCR), to tidy up the image before reading try Crop image, and to clean up stray spacing afterwards use Remove extra spaces.