We are using Realtime API. We are doing it with both VAD and Push to Talk. In both the scenario, even if AI gets interrupted, the recording/transcript shows the entire output that came out. Is there a way to exclude remaining part from the audio and transcript?
1 Like
I don’t know of a way to do that currently, The Realtime API is in Beta and a lot of changes are being made to improve things, transcription is one of those.
2 Likes
That shd be called out in the documentation then. Coz owise it misleads.