Hi guys, I’m working on a voice agent using VAPI and I’m having issues with high latency.
Even though the system latency shows around 575-1000ms, the actual response time during live calls is 3–4 seconds, and sometimes it goes up to 6–7 seconds.
My setup is really simple just a single prompt (314 words 1,935 chars) and no knowledge base or nothing.
I am attaching the audio file for one of my tests. I already tried reducing the prompt size, but the delay is still there.
Has anyone else experienced this? If yes, how did you fix it? What could be the reason? Any advice would be really helpful.
Thanks a lot in advance!