Technology
Meta Takes on OpenAI and Google With a Closed-Source Speech-to-Text Model
Muse Voice Transcribe challenges GPT Live Transcribe and Gemini Transcribe Live with lower latency and error rates.
Meta Platforms Inc. is entering the enterprise productivity arena with a proprietary real-time speech transcription model, a move that pits it directly against offerings from OpenAI and Google. Developed by Meta’s Superintelligence team, the new system, named Muse Voice Transcribe, is designed for instantaneous speech-to-text translation and speaker identification. The launch marks a notable deviation from the company’s standard practice of open-sourcing its major AI models.
