[sid-the-kid]

Thank you! Yes, right now we are using Qwen for the LLM. They also released a super fast TTS model that we have not tried yet, which is supposed to be very fast.