Thanks — you're asking whether the Ultravox plugin can be configured to return text-only output. I cannot confirm that Ultravox supports a

modalities=[ "text" ]

or equivalent setting. The Ultravox plugin docs list parameters like

voice

and

time_exceeded_message

but do not document a modality control. Would you like me to: • try configuring a realtime model + separate TTS (text-only model + a TTS plugin) as a workaround, or • attempt a quick test with your Ultravox setup if you can share a minimal config (plugin settings and how you instantiate it)? Sources: Realtime model plugins | LiveKit Docs | OpenAI Realtime API plugin guide | LiveKit docs | Gemini Live API plugin | LiveKit docs | Ultravox plugin guide | LiveKit docs