Voice to voice via API possible?

New to OpenAI and the API however I’m been trying to utilise 4o to code a python script however I’ve noticed the APi data in 4o on openAIs own documentation is way out of data and not at a v1 level.

This leaves me to ask here, is it even possible via the api to replicate the conversation type behaviour demonstrated in the release of 4o video? My assumption would be that this would need to be done with webrtc or similar for bidirectional interjections but I’m unable to find any reference in the documentation.

Short of capturing the apps packet flow and decrypting the SSL sessions to see what the app is doing I’m currently at a loss.

Any guidance?

No, not yet. Those capabilities aren’t available anywhere yet.