PixelRead: free, on-device OCR for anything on your Mac screen
I kept running into text I could see but couldn’t select—videos, images, PDFs, and remote desktops. I built PixelRead, a small macOS menu-bar app that turns any screen region into usable text.
Press ⌘⇧2, drag over a region, and a floating bar gives you Copy Text, Translate, and Speak. If text is already selectable, PixelRead can act on the selection without taking a screenshot. OCR and translation run on-device with Apple’s Vision and Translation frameworks, so captured text stays on the Mac. Recent captures are saved locally for quick reuse.
The core app is free: https://appunbound.com/app/pixelread
I’m looking for Mac users who regularly pull text from screenshots, video, PDFs, or apps: what would make this indispensable, and which workflows should I support next?
The on-device approach is the strongest part here. PixelRead isn't just extracting text from inaccessible surfaces; it keeps the workflow local, which gives the product a clear reason to exist beyond basic OCR.