Absolutely, you can turn off TTS (text-to-speech) and STT (speech-to-text) using RPC during a LiveKit session—it's pretty straightforward! If you want to disable STT, you just need to turn off the audio input with an RPC method like this:

@ctx.room.local_participant.register_rpc_method("toggle_input")
async def on_toggle_input(data: rtc.RpcInvocationData) -> None:
    if data.payload == "audio_off":
        session.input.set_audio_enabled(False)  # This stops the STT processing
    elif data.payload == "audio_on":
        session.input.set_audio_enabled(True)   # This turns STT back on

And if you’re looking to disable TTS, you can switch off the audio output similarly:

@ctx.room.local_participant.register_rpc_method("toggle_output")
async def on_toggle_output(data: rtc.RpcInvocationData) -> None:
    if data.payload == "audio_off":
        session.output.set_audio_enabled(False)  # This stops the TTS output
    elif data.payload == "audio_on":
        session.output.set_audio_enabled(True)   # This turns TTS back on

These handy methods give you the flexibility to control audio input and output on the fly during your session, making it easy to switch up how you interact. Sources: livekit-agents/livekit/agents/voice/agent_activity.py | Text and transcriptions | LiveKit Docs