← All tools

Voice agent latency budget

Everything between a caller finishing their sentence and hearing the first syllable back. Most teams blame the language model. It is usually the endpointing.

Start from a stack

What you get before anyone profiles it.

Response timePeople talk over it

1.51 s

At 1510 ms callers start speaking over the agent, which costs you a turn and usually a re-prompt. Endpointing alone is 40% of it.

The budget, as numbers

HopTimeShare
Network round trip140 ms9%
Endpointing (waiting for silence)600 ms40%
Speech-to-text finalisation200 ms13%
Routing / classification0 ms0%
Language model, first token450 ms30%
Speech synthesis, first audio120 ms8%
Total1510 ms100%

The point of this tool: load “Typical cascaded” and drag endpointing from 600 ms down to 250 ms. You just saved more time than switching to a faster language model would have, and it cost you nothing but a config value. Then compare the two routing presets — the same classification step costs 100 ms on a typed decision model and 400 ms on a chat model.

Why endpointing dominates

Every other hop is work. Endpointing is waiting. You are holding the line open, doing nothing, because you cannot yet prove the caller has finished their sentence. Set it short and the agent talks over people who paused to think; set it long and every single turn pays the penalty. Most SDKs ship a conservative default somewhere between 500 and 800 ms and most teams never touch it, which means the majority of their latency budget is spent on silence. Semantic turn detection — deciding from the words whether the thought is complete, not just from the pause — is the way out, and it is worth more than any model upgrade on this list.

What each band actually sounds like

Response timeWhat the caller experiences
Under 800 msIndistinguishable from a human on a slightly slow line. Nobody comments on it.
800 – 1,200 msA noticeable beat before every reply. Works, but the call feels stilted and long conversations tire people.
1.2 – 2 sCallers start filling the gap — “hello?” — and now you are handling barge-in on every turn.
Over 2 sAssumed to be a dropped call or a broken system. Hang-up rates climb sharply.

Assumptions

  • This measures time to first audio, not time to a complete answer. If you wait for the full response before speaking, add the whole generation.
  • Every hop assumes streaming. A single non-streaming stage invalidates the model — that is what the cross-region preset is really showing.
  • Speech-to-speech collapses transcription, reasoning and synthesis into one number under the language model. It wins on latency and loses on cost, control and per-layer failover, which is the trade covered in the cost calculator.
  • Network figures assume the media path terminates in the same region as your agent. Split those and the round trip doubles.

Related