META'S MUSE VOICE TRANSCRIBE

Summary from the TLDR AI newsletter

Muse Voice Transcribe is Meta's first real-time audio perception model. It supports streaming speech recognition, diarization for more than 20 speakers, endpointing, multilingual code-switching, and contextual biasing.