Voice¶

Reading answers aloud¶

• Read aloud reads an answer; Stop reading stops.
• Auto-read reads each answer when the work ends: the Whole answer, or a Short summary ("A sentence or two the Orchestrator adds to be said").
• How the Orchestrator speaks (up to 300 characters) is sent with your message when Auto-read reads a short summary.

Answers are read with this computer's voices, or with Gemini's:

• Gemini: add a Gemini API key and Save key. The key stays in KOLONİ; the page never sees it again. "When off, this computer's voices read."

Dictating¶

Dictate: "what you say is written into the box, and nothing is sent". Dictation language sets the language you speak; it also sets the language answers are read in.

Privacy¶

• "Speech recognition is your browser's: in Chrome and Edge what you say is sent to Google's or Microsoft's service to be written down."
• A computer voice marked online sends what it reads to its maker.
• "Text read with Gemini is sent to Google. On the free tier Google may use it to improve its products; on a paid key it does not."
• "KOLONİ keeps no sound and sends none anywhere."

Voice settings are kept in this browser: Settings › The app › Voice.