Meta Launches Muse Voice Transcribe: Real-Time Audio Perception Model Now Available
Key Info
Meta has released Muse Voice Transcribe, a real-time audio perception model from Meta Superintelligence Labs, now available through Meta Model API, Meta AI for Mac, and Muse Code.
Highlights
- Provides real-time streaming automatic speech recognition (ASR) and diarization for 20+ speakers.
- Uses adaptive delay to reach the Pareto frontier in the speed-accuracy trade-off, measured by time to final transcription.
- Available today across Meta's API, Mac app, and Muse Code tools.