I don't think that's exactly accurate. What's likely different is that GPT-4o can output the tonality instructions for text to speech now. It's probably the s

I don't think that's exactly accurate. What's likely different is that GPT-4o can output the tonality instructions for text to speech now. It's probably the same voice, but different instructions for generations. One was without tonal indicators, one with.