# What is latency in voice AI (and why it matters)?

> Latency in voice AI is the delay between the caller finishing speaking and the agent starting to reply. Under about one second feels natural; much more and the call feels laggy and the caller starts talking over the agent. Low latency is one of the biggest differences between a good and a bad voice agent.

Source: https://agents.elevatebox.in/glossary/latency  
Updated: 2026-10-12

## What drives latency

Listening, deciding the reply and speaking it each add delay, and so does the network and telephony route. India-based routing and providers tuned for Indian calls help keep the round trip short.

## How to measure it

Time the gap from the moment you stop speaking to the first sound of the reply, over several turns of a real phone call rather than a browser demo. The phone network adds delay a demo does not show, and the slowest turns are the ones callers remember.

## Frequently asked questions

### What is a good latency for a voice agent?

Replies that start within about a second feel natural on a phone call. Consistency matters as much as speed: a reply that is sometimes fast and sometimes slow feels worse than a steady one.

### Why is my AI caller slow to reply?

Each step adds time: recognising speech, deciding the reply, generating audio and the phone route itself. Long prompts, slow models and calls routed through faraway servers all add delay.