Hey, I’ve noticed similar latency spikes before and can help pinpoint whether it’s coming from GPT, 11Labs, or Retell.
Sometimes delays happen due to model load, API throttling, or network routing issues.
It would help to know if the lag is mostly on text generation, audio processing, or both.
With that info, we can isolate the root cause and suggest concrete fixes to bring your E2E latency back down.
@krustle_