Voxtral Mini

Open weights Deprecated Multimodal Dense voxtral-mini

3B open-weight Apache 2.0 speech understanding and transcription model.

Released
Jul 15, 2025
Context window
32.8K tokens
Max output
Not disclosed
Parameters
3B
Input price
Not disclosed
Output price
Not disclosed

Prices are first-party list prices in USD at the standard tier. Last verified Oct 30, 2026.

Pricing

Launch priceJul 15, 2025
Current price since Jul 15, 2025

Specifications

Developer
Mistral AI
Model type
Multimodal
Architecture
Dense
Parameters
3B
Input
Text Audio
Output
Text
Knowledge cutoff
Not disclosed
License
Apache License 2.0
API access
First-party API
Capabilities
Small LM