Voxtral-Mini-4B-Realtime available in Amazon SageMaker JumpStart
Mistral AI's Voxtral-Mini-4B-Realtime-2602 model is now available in Amazon SageMaker JumpStart for real-time speech transcription. This multilingual model enables low-latency audio-to-text applications and supports 13 languages with configurable delays. Customers can deploy it easily via SageMaker Studio or the Python SDK, allowing for quick integration into specific AI use cases.
- →Real-time speech transcription model available via SageMaker JumpStart
- →One-click deployment and SDK integration
- →Multilingual support and configurable latency
Features (2) ›
- Real-time speech transcription model available via SageMaker JumpStart
Amazon SageMaker JumpStart now offers Voxtral-Mini-4B-Realtime-2602, a multilingual, real-time speech-to-text model from Mistral AI. This model enables the creation of low-latency speech applications with a natively streaming architecture.
- One-click deployment and SDK integration
Customers can deploy the model with minimal effort through SageMaker Studio or programmatically via the SageMaker Python SDK, facilitating rapid integration into AWS accounts for specific AI use cases.
Enhancements (1) ›
- Multilingual support and configurable latency
The Voxtral-Mini-4B-Realtime model supports transcription across 13 languages and features configurable delays to balance latency and accuracy needs for various applications.
https://aws.amazon.com/about-aws/whats-new/2026/07/voxtral-mini-realtime-on-sagemaker-jumpstart/