I’m playing around with the transcription api via websockets, and I’m noticing that delta events arrive together with the final event once a pause is detected. This happens when VAD is both enabled and disabled. Is this the correct behaviour? Is it not possible to receive partials as the speaker is still speaking?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| GPT-4o-transcribe realtime, the .delta updates not received during the transcription | 7 | 483 | February 10, 2026 | |
| Realtime transcription messages flow is wrong | 15 | 1690 | August 8, 2025 | |
| Realtime streaming transcription | 4 | 247 | February 23, 2026 | |
| Discussion around syncing real-time AI-generated transcript deltas with WebRTC audio playback to ensure speech and on-screen text appear in natural alignment. | 3 | 387 | February 22, 2026 | |
| Input_audio_buffer.speech_stopped events not firing reliably with VAD (Realtime API) | 0 | 123 | November 30, 2025 |