Voxtral Small 24B 2507

Name: Voxtral Small 24B 2507
Author: Mistral

Mistral · Released Jul 15, 2025 · Open weights

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

Price per 1M tokens: $0.1 / $0.3
Context window: 32.8K

Lab announcement ↗Open in the timeline

BENCHMARKS

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

Mistral's past year

15 release days in the past year, about one every 24 days. The next is expected around Nov 16, 2026. Forecasts ↗

NovDecJanFebMarAprMayJunJulAugSepOctNov

More from the lab

Launch announcement · Feb 4, 2026Voxtral transcribes at the speed of sound.Launch announcement · Mar 23, 2026Speaking of Voxtral

Specifications

RELEASE DATEJul 15, 2025

WEIGHTSOpen weights

CONTEXT WINDOW32.8K

MAX OUTPUT32.8K

REASONINGNot listed

TOOL CALLINGNot listed

INPUTStext, audio/wav, audio/mpeg

OUTPUTStext

Models.dev record ↗

Specifications and prices from Models.dev; availability and limits may differ by provider. Benchmarks from Epoch AI (CC BY 4.0), LiveBench and the other sources named above.