Take a photo or screenshot of Urdu text and turn it into editable Urdu you can copy, correct and reuse. This Urdu OCR (optical character recognition) runs inside your browser, so your images are never uploaded.
What is Urdu OCR?
OCR stands for optical character recognition: software that looks at a picture of text and works out which letters it contains. For English, OCR has been reliable for years. Urdu is much harder, because its letters change shape depending on their position in a word, many letters differ only by dots, and the Nastaliq style stacks letters diagonally so that words overlap. This tool uses Tesseract, a well-known open-source OCR engine, with a model trained for Urdu, and adds its own clean-up step to fix common spacing mistakes.
How to convert an Urdu image to text
- Add your image. Choose a file, drag it onto the box, or copy a screenshot and paste it with Ctrl+V.
- Wait for the reading engine. The first time, your browser downloads the Urdu reading engine (about 5 MB). After that it is saved in your browser and starts almost instantly.
- Check and correct. The text appears in an editable box. Fix any mistakes directly, then copy it or download it as a .txt file.
How accurate is it?
Accuracy depends mostly on the style of the Urdu letters in your image.
- Naskh style (the upright style used on most websites, phones, subtitles and many printed forms) reads very accurately. In our tests on clear Naskh text, about 98 out of every 100 letters came out right.
- Nastaliq style (the flowing, slanted style of newspapers and books) is much harder for any OCR, because letters sit on top of each other and words overlap. In our tests about 93 out of every 100 letters were right, but spaces sometimes land in the wrong place. The tool repairs many of these spacing mistakes automatically using its Urdu word list, and you can fix the rest in the text box.
Tips for better results
- Keep it straight. Photograph the page from directly above, with the lines level.
- Use good light. Avoid shadows across the text and glare on glossy paper.
- Crop to the text. Remove pictures, borders and other columns. One column of text reads best.
- Screenshots beat photos. A screenshot of a website or PDF is sharper than a camera photo of a screen.
- Don’t zoom out too far. Text should be clearly readable to you in the image. Tiny text loses the dots that separate letters like ب ت ث.
What you can use it for
- Copying a quote from a book or newspaper without retyping it.
- Turning scanned documents, notices or old letters into editable text.
- Getting the text out of an image shared on WhatsApp or Facebook.
- Students collecting notes from printed material.
Frequently asked questions
Are my images uploaded to a server?
No. The reading engine runs inside your browser. Your image never leaves your device, which makes it safe for private documents.
Why is the first run slower?
Your browser downloads the reading engine and the Urdu language model once. They are cached, so the next image starts right away.
Can it read handwriting?
Not reliably. It is built for printed and on-screen text. Very neat handwriting sometimes works, but expect many mistakes.
Does it read English in the same image?
It is tuned for Urdu. English words mixed into Urdu text may come out wrong; numbers usually read fine.
Which file types are supported?
JPG, PNG, WebP and most other image formats your browser can open. For a PDF, take a screenshot of the page first.
Is it free?
Yes. There are no limits on the number of images.