To help you choose between responses from two LLMs and use TTS to speak the selected response in LiveKit, here’s a friendly step-by-step guide you can follow: 1. Generate Responses: You’ll want to use the

llm.chat()

method to get responses from both LLMs at the same time. 2. Implement Selection Logic: Then, apply your own criteria—like which response feels more relevant, higher quality, or quicker—to pick the best one. 3. Process Through TTS: Finally, send the chosen response to the TTS engine using the

performTTSInference

function or the

session.say()

method so it plays back in real time. Here’s a simple example to illustrate how you might set this up:

class MultiLLMAgent(Agent):
    def choose_response(self, resp1: str, resp2: str) -> str:
        return resp1 if len(resp1) < len(resp2) else resp2

    async def onUserTurnCompleted(self, chat_ctx: ChatContext, user_msg: ChatMessage):
        resp1_task = asyncio.create_task(self.llm.chat(chat_ctx=chat_ctx))
        resp2_task = asyncio.create_task(self.secondary_llm.chat(chat_ctx=chat_ctx))

        resp1, resp2 = await asyncio.gather(resp1_task, resp2_task)
        selected = self.choose_response(resp1.response, resp2.response)
        self.session.say(selected)

This way, you can tailor the selection process just how you like and smoothly bring in TTS for a natural voice output. If you have any questions or want to dive deeper into any part, I’m here to help—just let me know! Sources: examples/voice_agents/structured_output.py | Tool definition and use | LiveKit Docs | agents/src/voice/agent_activity.ts