Click to upload or drag & drop an image
PNG, JPG, WebP, BMP · English · processed locally
How does browser-based OCR work?
This tool runs Tesseract, the open-source OCR engine that powers countless scanning apps, compiled from C++ to WebAssembly so it executes directly in your browser tab. The first time you use it, your browser downloads the engine and the English recognition model (about 7 MB, cached for next time) — your image itself never leaves your device, which you can verify from the Network tab in DevTools.
What images work best
OCR accuracy depends far more on the input than the engine. Clear, printed text on a plain background — screenshots, scanned documents, book pages, receipts — recognizes very well. Small text, skewed photos, busy backgrounds, and decorative fonts degrade results, and handwriting is largely out of reach for any traditional OCR engine.
Frequently asked questions
Is my image uploaded to a server?
No. The OCR engine runs as WebAssembly inside your browser, so recognition happens on your own CPU. The only network activity is a one-time download of the engine and language model — open DevTools and watch the Network tab while you process an image if you'd like to confirm.
Why is the first run slower than later ones?
On first use your browser fetches the WebAssembly engine and the English model (about 7 MB combined) and caches them. After that, the engine starts almost instantly, and recognition speed depends only on your device and the image size.
Can it read handwriting?
Not reliably. Tesseract is trained on printed typefaces. Neat block capitals sometimes partially work, but cursive or ordinary handwriting produces mostly noise — that task needs specialized handwriting-recognition models.
Which languages are supported?
English, currently. Each additional language needs its own trained model file of a few megabytes, so we started with the most requested one. If you need another language, email us — adding one is straightforward.
Why did my scanned PDF not work here?
This tool reads image files (PNG, JPG, WebP, BMP). For a scanned PDF, first export the page as an image — our PDF to JPG tool does that in the browser too — then run the image through this tool.