Skip to content

Fact · Speech recognition

Deepgram defines end-of-turn latency as the time from when a user stops speaking to when an end-of-turn event is received.

Checked against its source on October 17, 2026

In the source's own words

“The time from when a user stops speaking to when an EOT event is received.”

Measuring STT latency (Deepgram docs), Deepgram, Deepgram, n.d., Latency metrics.

Open the original
What it measures
Deepgram says this is the critical metric for voice agents.
What it does not show
A definition from one vendor. The end-of-turn event is Deepgram's own signal, so the number depends on its endpointing.

Source

Cited on

Deepgram defines end-of-turn latency from the end of speech | Yap