Translate the text inside images, right in your browser
The Caliptic Translate browser extension translates what is on a page — including the words inside images — and reads it aloud. Recognition and translation both run on your own computer.
Three modes, for three different problems
Text mode reads the page itself: select a passage or click a paragraph, and it is translated and read aloud. It works on sites that block text selection, because it reads the text from the page rather than through the selection. Image mode takes a picture of the region you choose and recognizes the words inside it — a menu photograph, a sign, a screenshot, an infographic. Live mode asks nothing of you at all: whatever is on screen is read, the translation is drawn over the original text, and each block is spoken as it scrolls into view.
Nothing you read leaves your computer
Translation uses the browser's own built-in, on-device model. Speech uses a local system voice — the extension explicitly skips network-backed voices. Text recognition runs through a small helper that ships inside the Caliptic Translate desktop app and communicates over a local channel, not the network. The extension also has no always-on presence: it declares no content scripts and no host permissions, so it stays inert until you click it. That is not a promise about our intentions; it is visible in the permission list before you install.
What it needs, and what it does not
Text mode works on its own, on any platform, with nothing else installed. Image and live modes need the free Caliptic Translate app for macOS, because browsers have no text recognition of their own; install the app, open it once, and it sets up the browser connection itself. On Windows and Linux the extension tells you this instead of failing silently. There is no account, no advertising and no tracking, and the extension is free.
Does it work on pages that block copying and selecting?
Yes. Instead of going through the selection, you click the paragraph you want and the extension reads the text from the page itself — the same place a browser's reading mode and a screen reader read it from. It is meant to help you understand the page you are on; it is not a tool for bulk copying, and the site's terms remain your responsibility.
Why does translating images need a desktop app?
Browsers cannot recognize text inside an image; there is no such capability in the browser platform. The alternative would be uploading your screen to a server, which is exactly what we did not want to build. Instead the recognition happens on your own machine through Apple Vision, inside the app you already installed.
See also
Translate webtoons and comics as you scroll · What leaves your device (and what never does) · Real-Time Translator for iPhone · Real-Time Translator for Android · Translate My Call: phone calls answered on your computer · A translator device alternative you already own · Translate With One Earbud Each · Offline translator app for iPhone and Android · Travel translator app that works with no signal · Caliptic vs Timekettle: earbuds you already own · Caliptic vs Pocketalk: no second device, no data plan · Caliptic vs Vasco: more languages, or no connection needed · Translate Computer Audio in Real Time · On-Device Real-Time Translation · Real-Time Translation for Windows · Real-Time Translation for Mac · Translate Meetings in Real Time · On-Device Meeting Notes · Frequently Asked Questions