Voxtral Transcribe 2 consists of two next-generation speech-to-text models with state-of-the-art transcription quality, diarization, and ultra-low latency. The family includes Voxtral Mini Transcribe V2 for batch transcription and Voxtral Realtime for live applications.

Voxtral Mini Transcribe V2 offers state-of-the-art transcription with speaker diarization, context biasing, and word-level timestamps in 13 languages. It maintains noise robustness in challenging acoustic environments and supports processing recordings up to 3 hours in a single request. Voxtral Realtime is purpose-built for live transcription with latency configurable down to sub-200ms, enabling voice agents and real-time applications.

Voxtral Realtime uses a novel streaming architecture that transcribes audio as it arrives, unlike approaches that adapt offline models by processing audio in chunks. The model delivers transcriptions with delay configurable down to sub-200ms, unlocking a new class of voice-first applications.

Voxtral powers voice workflows in diverse applications including meeting intelligence, voice agents and virtual assistants, contact center automation, media and broadcast, and compliance and documentation. Both models support GDPR and HIPAA-compliant deployments through secure on-premise or private cloud setups.

The models are designed for developers building voice applications, enterprises needing transcription services, and organizations requiring privacy-first deployments. Voxtral Realtime ships under Apache 2.0 license for deployable on edge applications, and both models integrate with existing systems through API access.