Meta releases Muse Voice Transcribe with top streaming accuracy

ArtificialAnlys · x · 2026-09-02

Meta Superintelligence Labs released Muse Voice Transcribe, the first streaming Speech-to-Text model. It achieves #1 Final Transcript accuracy on AA-WER Streaming with 3.1% WER at 0.16s latency, supporting 70+ languages. Priced at $0.18/hour, it undercuts competitors like Cartesia Ink-2 and ElevenLabs.

Related event: Meta Launches Muse Voice Transcribe, Claiming SOTA Streaming ASR(10 posts)→

Original post →

More from Multimodal

Multimodal channel →