So far sdialog.audio only supports Kokoro or Hugging Face TTS. It would be good to allow support to more powerful TTS like IndexTTS.