You photograph a page — a book, magazine, or article. The photo is sent to Google's Gemini model, which transcribes the printed text into structured text (headings, paragraphs, verse, captions) in the order it should be read. Your device's own text-to-speech engine then reads that text aloud, at the speed and in the voice you choose.
The photo is processed to produce the transcription and is not kept longer than necessary to do that. A record of the read (outcome, page and character counts, and — only if enabled — the transcribed text) may be stored for debugging and to improve accuracy; it is not linked to any account, because Vachak doesn't have accounts.
It doesn't translate the text by default — it reads back the page's own language, using whatever voice your device has installed for it. It doesn't build a saved library of what you've photographed. It doesn't require a sign-in.
Photographing a page you own, for your own reading, is squarely within ordinary personal use. Vachak is a reading tool for your own material — please use it that way.