IconVault's OCR Extractor pulls text out of images using on-device OCR powered by Tesseract.js, so your image is never uploaded to a server. Drop in a screenshot, photo or scan, watch the live progress bar, then copy the recognized text or download it as a .txt file. Recognition happens right in your browser tab.

It is free with no signup and fully private. Two honest notes: the first run downloads the English language model once (it is cached after that), and accuracy depends on image quality, so sharp, high-contrast images give the best results. Stylized fonts, heavy blur or busy backgrounds will reduce accuracy.