Absolutely! An agent can indeed speak while running function tools in parallel. This happens when the

parallel_tool_calls

option is turned on during the agent's setup. How the agent talks depends on whether the tools send back any output: • If at least one tool gives a result, the agent will kindly wait and then respond once all tools have finished their work. • If you prefer the agent to stay quiet after the tools run, you can simply use

ev.cancel_tool_reply()

within the

function_tools_executed

event handler. If you'd like, I’d be happy to share a detailed code example to help you see exactly how to set this up. Just let me know! Sources: xAI LLM integration guide | LiveKit Docs | xAI LLM integration guide | LiveKit Docs | Perplexity LLM integration guide | LiveKit Docs | Cerebras LLM integration guide | LiveKit Docs | livekit-agents/livekit/agents/voice/generation.py