Read Arabic script from an image, with right-to-left reading order preserved in the output rather than reversed. Arabic letters change shape depending on their position in a word, which is what makes the script hard for OCR and what this pack is trained for. The same model reads Persian, Urdu, Uyghur, Pashto, Kurdish, Sindhi and Balochi.
How it works
- Drop an image or PDF. Text detection runs immediately.
- Pick your language pack (downloaded once, 8–17 MB) and press Read text.
- Copy the lines, or download .txt / .json. PDFs are read cover to cover.
Questions
- Is the text right-to-left in the output?
- Yes. Reading order comes from the language pack rather than from guessing at the characters, so Latin words inside an Arabic line are left alone instead of being flipped. The searchable-PDF export writes the same order.
- Does my file get uploaded?
- No. The models and all processing run inside your browser on your own device. You can verify this yourself: open your browser’s DevTools, watch the Network tab, and process a file: there are no uploads. After the first visit most tools also work fully offline.
- Which languages are supported?
- 106 languages across Latin, Cyrillic, Arabic, Devanagari, CJK, Greek, Thai, Tamil and Telugu scripts. Bengali is the one major gap, since no v5 model exists for it yet, and the picker says so rather than guessing.
- How accurate is it?
- On clean document text, the measured character error rate is under 2%, and every line carries a confidence score so you can see which ones to check. Handwriting and low-resolution photos will read worse.
- Does it read scanned PDFs?
- Yes. Digital PDFs are read from their own text layer, with exact characters at exact positions. Scanned pages go through detection and recognition, and mixed documents use both.