mistralaiListed Jul 19, 2025

Voxtral-Mini-3B-2507

Voxtral Mini is an enhancement of Ministral 3B, incorporating state-of-the-art audio input capabilities while retaining best-in-class text performance. It excels at speech transcription, translation and audio understanding.

Context window32.8K
Parameters3B
Provider endpoints2
Benchmark results9
Measured performance

Benchmark results

All leaderboards ↗

Word error rate · lower is better

Open ASR · AMI-Cleaned WER ↗English short-form; normalization unversioned
13.57% WER

mistralai/Voxtral-Mini-3B-2507

Leaderboard entry
Open ASR · LS Clean WER ↗English short-form; normalization unversioned
1.48% WER

mistralai/Voxtral-Mini-3B-2507

Leaderboard entry
Open ASR · LS Other WER ↗English short-form; normalization unversioned
3.63% WER

mistralai/Voxtral-Mini-3B-2507

Leaderboard entry
Open ASR · SPGISpeech WER ↗English short-form; normalization unversioned
2.15% WER

mistralai/Voxtral-Mini-3B-2507

Leaderboard entry
Open ASR · URGENT2024 WER ↗English short-form; normalization unversioned
15.66% WER

mistralai/Voxtral-Mini-3B-2507

Leaderboard entry
Access & cost

Prices across providers

2 quotes

DeepInfra

Input audio second

Rate
$0.0000166667
USD / second
32.8K contextPrice source ↗Checked Oct 17, 2026
Endpoint IDmistralai/Voxtral-Mini-3B-2507

Model specifications

Developer
mistralai
Modality
Audio
Context
32,768 tokens
Maximum output
Not reported
Weights
Open weights
Provider listed
Jul 19, 2025
First discovered
Oct 16, 2026

Exact identities

  • openroutermistralai/voxtral-mini-3b-2507
  • deepinframistralai/Voxtral-Mini-3B-2507
  • open-asrmistralai%2FVoxtral-Mini-3B-2507

Sources

Updated Oct 17, 2026

More from mistralai

View all ↗
mistralaiListed Apr 19

Mistral: Voxtral Mini TTS

Voxtral Mini TTS is Mistral's text-to-speech model featuring zero-shot voice cloning and multilingual support. It converts text input into natural-sounding audio output.

4.1K contextNo reported score
No reported score
Price not listed