Image to Text (OCR) reads letters from a picture. Clear pictures give clean text; blurry or tilted ones give mistakes or No text found.
Get better results
- Pick the right language in Language in the image (English, Hindi, Punjabi, Korean, Japanese, Chinese and more) before tapping Read text.
- Light: photograph pages in good, even light without shadows.
- Straight: hold the camera flat over the page, not at an angle.
- Sharp: tap to focus; screenshots work better than photos of screens.
- Crop to just the text you need.
It is slow the first time
The first time you use a language, the tool downloads that language’s data. After that it is much faster.
Handwriting
The tool is made for printed text. Neat handwriting may partly work, but expect mistakes.
Mixed languages
For documents with both English and Hindi (common on Indian forms and bills), choose English + Hindi. Choosing only one language makes the other come out as nonsense.
Reading the confidence
When it finishes, the tool shows Done! Confidence with a percentage. Above about 85% the text is usually clean; below 60%, retake the photo with better light or crop closer.
Tips for common sources
- Screenshots: crop away menus and buttons so only the text is read.
- Old or faded paper: increase contrast first with Photo Filters and Adjustments (black & white plus contrast).
- Scanned PDFs: convert the pages with PDF to Image and read each page.
- Tables: OCR reads row by row as plain text; check numbers carefully before using them.
Tool: Image to Text (OCR)
Still stuck?
Contact support. Tell us which tool or page you were using, what you tapped, what you saw, and your phone or computer and browser (for example: Android phone, Chrome). We reply by email.
Questions, or an error not listed here?
Tell us in the comments: what you were doing, the exact message you saw, and your device and browser. We read every comment and add new fixes to this article.