Voxtral Mini 3B 2507

Open audio-language model for speech transcription, audio understanding, and voice-driven tool use

BENCHMARKS

Not in the Epoch Capabilities Index. Epoch benchmarks what it can run, so this is a gap in coverage, not a poor result.

From the lab

Launch announcement · Jul 15, 2025VoxtralLaunch announcement · Feb 4, 2026Voxtral transcribes at the speed of sound.Launch announcement · Mar 23, 2026Speaking of Voxtral
RELEASE DATEJul 15, 2025
WEIGHTSOpen weights
CONTEXT WINDOW32.8K
MAX OUTPUT32.8K
REASONINGNo
TOOL CALLINGYes
INPUTStext, audio
OUTPUTStext

Specifications from Models.dev. Availability and limits may differ by provider. Benchmark scores come from Epoch AI under CC BY 4.0.