Thanks — I’m responding to your question about whether LiveKit supports context‑aware TTS. Short answer: LiveKit does not provide autonomous context‑aware TTS (a TTS model that dynamically adapts prosody/intonation from conversation context). What you can do in LiveKit: • Use speech tags (e.g.,

[laugh]

,

[pause]

,

<whisper>

) with xAI TTS to control delivery programmatically. • Use detailed performance prompts with Gemini TTS to request a style (e.g., “warm and unhurried”); this requires prompt engineering rather than automatic context adaptation. • Integrate other TTS plugins (ElevenLabs, Polly, Inworld) and drive style via parameters/tags you insert. Note: LiveKit’s context‑aware turn detector improves end‑of‑utterance detection and turn-taking (helps when mixing STT/VAD), but that doesn’t make TTS generation itself context‑aware. Would you like examples of speech tags or a sample Gemini prompt to achieve a specific speaking style? Sources: xAI TTS | LiveKit Documentation | A Practical Guide to Prompting Gemini 3.1 Flash TTS | LiveKit | LiveKit turn detector plugin | LiveKit Documentation, with interactive elements