How to get text from an image
- Choose the language of the text: English, Hindi or both.
- Add a photo, screenshot or scanned PDF.
- Wait while it reads. A progress bar shows each step.
- Copy or save the text. You can correct anything in the text box first.
Tips for better results
Photograph the page straight from above in good, even light, and fill the frame with the text. Avoid shadows and glare. If a photo is sideways, fix it with Rotate & Flip Image first, and crop away the background with Crop Image. Higher resolution helps with small print.
Image to text is useful for copying a paragraph from a book, digitising old letters and certificates, grabbing text from screenshots, and turning scanned documents into something you can search and edit. Because it runs on your device, it's safe for private papers such as ID cards and medical reports.
Students use it to copy questions from textbook photos into notes, office staff to pull figures from printed invoices, and anyone filling a form to copy long numbers from a card without typing mistakes.
Questions
- Which languages are supported?
- English and Hindi, or both together for mixed documents. More languages may be added later.
- How accurate is it?
- Clear, straight photos of printed text are usually read very accurately. Handwriting, blurry photos, unusual fonts and very small text are harder. Always check names, numbers and amounts.
- Can it read scanned PDFs?
- Yes. Add the PDF and up to 20 pages are read one after another. For text PDFs made on a computer, PDF to Text is faster and exact.
- Why is the first run slow?
- The text reader (about 7 MB with English) is downloaded from this site the first time and then cached by your browser, so later runs start quickly.
- Are my images uploaded?
- No. Text recognition runs inside your browser using the open-source Tesseract engine. Your images and documents never leave your device.