Mistral AI·2026-02-05 00:00· 2026-02-05

Voxtral transcribes at the speed of sound.

AI summary

Mistral released Voxtral Transcribe 2, comprising Voxtral Mini Transcribe V2 for batch transcription and Voxtral Realtime for real-time use.

Selection record

Threshold 60Official, first-handFirst 71Second 74

AdmittedSum of both 145 ≥ twice the threshold 120

Source tier: Official, first-hand; this tier's threshold is 60
Pre-filter: passed：发布语音转文本AI模型及API
Why it was chosen: Latency, error-rate and pricing comparisons clarify options for real-time speech applications.

A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.

Source: Mistral AI · mistral.ai

#Model releases #Speech #Open-source ecosystem #Hugging Face