OCR — optical character recognition — is the technology that turns a picture of text into actual, selectable, searchable text. It's what makes a scanned contract searchable, lets you copy a quote from a photo, and powers the "search your scans" feature in modern document apps.
How OCR works in 2026
Modern OCR uses neural networks trained on millions of pages of real-world text. The accuracy on clean English documents is now over 99%, and even handwritten text is finally usable for many languages.
Picking the right OCR for your language
- English documents: use the OCR English tool — fastest, highest accuracy.
- Arabic, Hebrew, or right-to-left scripts: use OCR Arabic (purpose-tuned).
- Mixed-language PDFs: use OCR Multi-language — auto-detects 100+ languages.
OCR best practices
- Scan at 300 DPI or higher when possible.
- Straighten skewed pages before OCR.
- Crop out shadows and noisy backgrounds.
- For multi-page PDFs, use OCR PDF instead of running each page separately.
After OCR: convert to Word, Excel, or plain text
Once your scan is OCR'd, you can convert it to a fully editable Word document, extract tables to Excel, or pull the raw text out as a .txt file. All from the same toolkit.