latency
# supportt
thenumu
04/18/2024, 5:41 PMHi can you confirm if the latency app on the dashboard is correct?
Every latency is around 1000ms minimum and 1300 for GPT4. This is around 3 times published figures for deepgram/groq
No functions just basic prompts..you can see this with your dashboard examples loaded
Also if you add twillio that instantly doubles it to around 2000m
Where are the Vapi servers located and do they geo load balance?
Thanks
v
Vapi Support
04/18/2024, 5:41 PMHi @thenumu, I'm an AI assistant for Vapi. While a team member reviews this question, I'll look through relevant sources and see if I can provide suggestions. Please tag @Vapi Support with any follow-up questions.
Vapi Support
04/18/2024, 5:41 PM### Answer
Vapi aims to provide a superior conversational experience and has implemented various latency optimizations like end-to-end streaming and colocating servers to reduce latency as much as possible [(1)]() . It manages the coordination of interruptions, turn-taking, and other conversational dynamics to ensure responsiveness. However, the specific latency figures you mentioned are not directly addressed in the information provided. Regarding server locations and geo load balancing, this detailed technical setup information, including whether Vapi employs geo load balancing, is not specified in the available data. For further assistance and more detailed technical insights, reaching out directly to Vapi's support might provide you with precise information.
Vapi Support
04/18/2024, 5:42 PMGot it. Tagging @User for additional help.
9 Views