Drop a screenshot or photo. Everything happens on your own device — nothing is uploaded anywhere.
Drop an image, or click to choose one
PNG, JPG, WEBP — works best on clean screenshots and printed text
Extracted text
Nothing extracted yet.
Upload an image on the left. The first image ever processed may take a little longer, since the reading engine downloads once and is then cached by your browser. After that it's quick.
About the sticky panel: it stays pinned in this page as you scroll, so it's easy to get back to. It can't float over a different website or tab — no page can reach outside its own browser tab. Copy the text here, switch tabs, and paste it in.
About accuracy: this runs a free, open-source OCR engine (Tesseract) entirely on your device. It's solid on clean screenshots and printed pages, but weaker on handwriting, low-contrast photos, or busy backgrounds. If you're extracting from tricky images regularly, an AI vision model would do noticeably better — just at a small per-image cost.