Recognize and extract text from photos, scans, whiteboards, and screenshots. Runs on-device with zero server uploads.
Docsworker brings neural-network OCR directly into your web browser using WebAssembly. When you scan an invoice, book page, or business card, the character recognition executes on your device's graphics and CPU processors. Your images never leave your local hardware, making this safe for corporate data, patient records, and personal receipts.
Image to Text uses optical character recognition (OCR) to read the words in a photo, scan, whiteboard picture or screenshot and give them back as editable text. It accepts JPG, PNG and WebP images and supports English, Urdu, Arabic, Hindi, French, Spanish and Simplified Chinese.
An Auto-Enhance option boosts contrast and sharpens edges, which helps with dim or slightly blurry photos. When the result appears you can correct it in the text box, copy it to the clipboard or download it as a .txt file.
Results are best with well-lit, in-focus pictures where the text is straight and clearly visible. The language data is downloaded once and kept in your browser cache, while the image itself is processed on your device and not uploaded.
English, Urdu, Arabic, Hindi, French, Spanish and Simplified Chinese. Choose the language that matches the text in your picture.
Blur, shadows, low resolution or unusual fonts reduce accuracy. Try a sharper photo or switch on Auto-Enhance, then fix any remaining mistakes in the text box.
No. Recognition runs in your browser. Only the language data is fetched the first time, and your image stays on your device.