There is something from the demo that still naws at me. Mira said, that the 4o model reasons across voice (audio), text, and vision (video). I still don't see any indicators of this for api usage and consumption whatsoever. Firstly, I am asking is this a model consolidations from an api perspective for creators or is this something internal availability only for ChatGPT-4o itself? I will use audio and video as examples. Text has come with an iterative stream feature so this is the kind of feature set I am looking for that correlates with the demo and it's output capabilities. Audio Audio falls under Speech-to-Text (STT) and Text-to-Speech (TTS). In the case of this concern we are speaking to the 'whisper model' modality via the api docs and more specifically STT because that would be the i...
No Comments 
