Not established from the available sources.
MISTRAL · MODEL RELEASE TRACKER
Voxtral Small
Voxtral Small is an open speech/ASR model (part of the Voxtral family) reported in comparisons as the top open model with a low word error rate (~2.77%).CURRENT SNAPSHOT3/5 DIMENSIONS WITH DATA
The dimensions that change the decision.
Not established from the available sources.
ModalityAutomatic speech recognition / speech transcription (audio ASR).
Reported word error rateReported word error rate ≈ 2.77% in cited ASR comparisons (identified as the best open model in that comparison).
AvailabilityDescribed in comparisons as an open model (open-weight / open).
RELEASE TIMELINE
Published, source-backed release events only.
VERIFIABLE FACTS
Every value stays attached to a source and date.
MODALITIES · ModalityDEVELOPER CLAIM
Automatic speech recognition / speech transcription (audio ASR).
AVAILABILITY · AvailabilityDEVELOPER CLAIM
Described in comparisons as an open model (open-weight / open).
BENCHMARKS · Reported word error rateDEVELOPER CLAIM
Reported word error rate ≈ 2.77% in cited ASR comparisons (identified as the best open model in that comparison).
WHAT CHANGED