Image to Text (OCR)
Turn a photo or screenshot of printed Amharic or English into text you can copy and edit. Pages that mix both languages, with dates and numbers, are read too. Everything runs on your device.
- Never uploaded
- Free, no sign-up
- No watermark
- Works on phones
- Works offline after one download
Your images are read on your device and never uploaded.
Pages
The first time, the reader and its language models download once (about 6 MB for Amharic + English) and are then kept on this device.
Text
yellow the reader wasn’t sure · green corrected with the Amharic word list (tap it to switch back to what was read). You can edit the text directly.
Works offline after one download.
See what it does
Before: Photo of a page
After: Text it foundHow to copy Amharic or English text from a photo
- Add photos or screenshots of the pages, or take a photo of a page. Rotate any page that is sideways.
- Keep “Amharic + English” selected and tap “Read text”. Phone photos are cleaned up first to remove shadows.
- Check the highlighted words, edit anything that’s wrong, then copy the text or download it as a .txt file.
Why use Picsawa
- Private by designFiles are processed on your device and never uploaded.
- Free, no sign-upNo account, no watermark and no daily limits.
- Works everywherePhones, tablets and computers, in any modern browser.
About the Image to Text (Amharic & English) tool
The reader uses Tesseract’s Amharic and English models, running in your browser. After reading, words that look like English letters inside Amharic lines are read again with the Amharic model, and words that are not Amharic words are compared with a list of 66,000 common Amharic words to fix typical misreadings such as ን/ነ, ስ/ሰ and አ/እ.
In our test on ten kinds of printed Amharic text (letters, notices, news, forms, proverbs, mixed Amharic and English) in four common fonts, these fixes cut the number of wrong words by about a third. Clear fonts such as Noto Sans Ethiopic came out almost perfectly; decorative or serif fonts and very small screenshot text need more checking.
Your images are never uploaded. The engine and models (about 6 MB) are downloaded once and stay in your browser. The Amharic word list is built from Amharic Wikipedia (CC BY-SA 4.0).
Frequently asked questions
How accurate is the Amharic text?
On clear printed text, about 19 of every 20 letters are read correctly in our tests, and much better with clear fonts such as Noto Sans Ethiopic. Small screenshot text, decorative fonts and dark or blurry photos need more checking, which is why unsure words are highlighted.
Are my images uploaded?
No. The text is read in your browser with an on-device engine. Your photos and the text never leave your phone or computer, so private letters and forms stay private.
Can it read handwriting?
No. It reads printed or typed text: letters, books, forms, notices, screenshots. Handwriting usually comes out wrong.
How do I get the best result?
Photograph the page flat and straight, in good even light, close enough that the text fills the photo. Rotate pages upright before reading. Screenshots of clear text work very well.
Why does the first read take longer?
The first time, the reading engine and language models are downloaded once (about 6 MB for Amharic + English). They are then kept in your browser, so the next time starts right away, even offline.
What do the highlighted words mean?
Yellow words are ones the reader wasn’t sure about, so please check them. Green words were corrected using a list of common Amharic words; tap one to switch back to what was read. You can edit all of the text before you copy it.