Absolutely! You can create a

@function_tool

that sends just text—no TTS involved—to the client in LiveKit. Let me share an example of how you might set up an Agent class specifically for text-only communication: Example Agent Class for Text-Only Output:

class MyAgent(Agent):
    def __init__(self) -> None:
        super().__init__(
            instructions="You are a helpful assistant.",
        )

    @function_tool
    async def get_weather(self, location: str):
        """Called when the user asks about the weather.
        
        Args:
            location: The location to get the weather for
        """
        return f"The weather in {location} is sunny today."

Configuration for Text-Only Communication:

async def entrypoint(ctx: JobContext):
    session = AgentSession(
        llm=openai.LLM(model="gpt-4o-mini"),
        # Notice we don’t need TTS or STT here
    )
    await session.start(
        agent=MyAgent(),
        room=ctx.room,
        room_input_options=RoomInputOptions(text_enabled=True, audio_enabled=False),
        room_output_options=RoomOutputOptions(transcription_enabled=True, audio_enabled=False),
    )

Key Points: • Text-Only Configuration: We set

room_input_options

and

room_output_options

to enable text input and disable audio, keeping things simple. • No TTS Needed: Since audio output is off, the Agent doesn’t need a TTS instance. • Function Tool: The

@function_tool

decorator helps the agent handle specific tasks and return text responses without any TTS. This way, your agent sends only text to the client, making it perfect for text-only interactions. If you have any questions or want to explore this more, just let me know!