Meta Introduces Muse Voice Transcribe, a Real-Time Speech-to-Text Model

Meta launched Muse Voice Transcribe, an autoregressive multimodal language model from the Muse Spark family designed for real-time streaming speech recognition. The model supports diarization with more than 20 speakers, endpointing and multilingual recognition with code-switching, making it suitable for meetings, call centers and media transcription. Muse Voice Transcribe is available through the Meta Model API, Meta AI for Mac and Muse Code, extending Meta's push into developer-facing AI tools alongside its larger Muse Spark models.