plus i dont have the range of voices since it seems to use the realtime ai voices from Openai which makes sense