Meta released 'Muse Voice Transcribe', an AI-powered speech-to-text model providing real-time transcription capabilities, according to a Meta official product announcement on September 2, 2026. The tool currently supports five Indian languages, aiming to facilitate multilingual communication, per Meta's technical release documentation.
According to a Meta spokesperson, 'Muse Voice Transcribe is designed to bridge the communication gap in diverse linguistic markets by providing high-accuracy, real-time transcription.' The specific list of the five supported Indian languages and the exact latency metrics for transcription remain unverified in the initial release notes.
This follows the release of OpenAI's Whisper model, which also provides multilingual transcription, though Meta's focus is specifically optimized for Indian regional dialects, per comparative technology reviews. Indian content creators and regional media organizations can now automate transcription workflows, potentially reducing production costs for multilingual video content, according to an assessment of regional digital media markets.
Privacy advocates, including representatives from digital rights monitoring groups cited in tech policy disclosures, have raised concerns regarding the data processing of voice inputs, citing potential risks in Meta's handling of regional language datasets.
