Voxtral transcribes

Voxtral Transcribe 2 is a new family of speech-to-text models delivering state-of-the-art transcription quality with precision diarization and ultra-low latency. The family includes Voxtral Mini Transcribe V2 for cost-effective batch transcription and the open-weights Voxtral Realtime, which is purpose-built for live applications like voice agents. Key enterprise features include identifying who spoke when, biasing the model for technical terms, and robust support for 13 languages. These models provide industry-leading accuracy at a significantly lower cost, outperforming major competitors while processing audio much faster.