Omi
omi-medical-1
Medical English, served from Ohio
Measured 2026-10-16
Build voice agents on Speko →omi-medical-1 is measured here for comparison — it is not one of the models Speko routes.
WER · batch
16.67%
95% CI 8.1–26.6 · n=50 · vendor-direct batch · FLEURS
The batch API returned nothing on 7 of the 50 clips, the corpus’s quietest, scored as all-deletion. On the other 43 it reads 3.1%.
WER · streaming
Read speech
4.04%
Conversation
15.01%
Details
Read speech
n=1541 · FLEURS test+dev+train · vendor-direct
Conversation
12.8–17.9% · EdAcc n=283 · 60 speakers, 2 passes, 2026-10-16
Finalize
722ms
p50–p90 · US East · n=200 · vendor-direct · re-measured 2026-10-20 · ±7%
Measured 2026-10-20
Marker-limited: 54 of 200 readings land before the measured speech end, 26 by more than 0.5s (Modulate on the same clips: 26 and 12).
Time to first token
962ms
p50–p90 · US East · n=200 · vendor-direct · re-measured 2026-10-20 · ±10%
Measured 2026-10-20
First words come from a separate draft model; omi-medical-1 writes only the final.
Cost / min
$0.0075
Vendor list, $0.45 per live audio hour.
Share this result
Embed the live badge
[](https://speko.ai/benchmarks/stt/omi-omi-medical-1)
<a href="https://speko.ai/benchmarks/stt/omi-omi-medical-1"><img src="https://speko.ai/benchmarks/badge/stt/omi-omi-medical-1.svg" alt="Speko stt rank"></a>
https://speko.ai/benchmarks/stt/omi-omi-medical-1