Media
3 items
Description
Right-click any image to select and copy the text inside it. OCR runs in your browser — images never leave your device. Screenshots are where text goes to die. Page Text brings it back. SELECT AND COPY Right-click any image and choose "Select text in this image". A transparent, selectable text layer appears over the picture — drag to select, copy with Cmd/Ctrl+C, or hit "Copy all". Works on screenshots, photos of documents, slides, error messages, anything with text in it. FIND INSIDE IMAGES Press Alt+Shift+F. Your browser's own find bar stops at the edge of an image; this one doesn't. Type a word and every image on the page containing it lights up, with the matching line highlighted. Images are read in the background as you scroll, so search is instant once a page is indexed. JAPANESE AND ENGLISH PP-OCRv5 reads both, including mixed lines. 100% ON-DEVICE There is no server and no API key. The OCR models (~43 MB) are downloaded once from Hugging Face, cached, and then run locally: text detection on WebGPU, recognition on multithreaded WebAssembly, via LiteRT.js. Your images are never uploaded — there is nowhere for them to go. - Works offline after the one-time model download - Free, no account, no tracking, no analytics - Results are cached per image, so revisiting a page costs nothing - Open source: https://github.com/john-rocky/page-voice