Wispr Flow's Hinglish bet pays off in India voice AI market

Meta has launched Muse Voice Transcribe, a real-time speech-to-text model priced at $0.18 per hour of audio that combines transcription, speaker diarization for 20+ speakers, and endpoint detection in a single model. The system supports over 70 languages, with 25 extensively validated for release, and handles multilingual code-switching without separate post-processing. While competitors like Speechmatics support higher speaker counts (up to 100), Meta's combination of low latency, high-capacity diarization, and aggressive pricing targets enterprise developers building meeting systems, call analytics, and ambient AI applications.

Meta unveiled Muse Voice Transcribe, a new audio AI model that transcribes speech to text and segments audio by speaker. CEO Mark Zuckerberg announced the model on Threads, noting it was trained on more than 70 hours of audio data. The model represents Meta's continued push into audio AI capabilities alongside its existing generative AI portfolio.

Clipto, a three-year-old AI startup that uses machine learning to search large video datasets, has reached a $250 million valuation after raising $15 million in new funding. The company achieved $15 million in annual recurring revenue and profitability before closing the round. The funding reflects investor confidence in AI-powered video search as a commercial tool for handling terabytes of media.

Google has released Gemini 3.5 Transcribe, a new transcription model that automatically detects specialized jargon and supports more than 85 languages. The model represents an improvement over its predecessor, Chirp 3, with better multilingual performance and lower wording error rates. Users can edit transcriptions using voice commands. The release comes as Google continues to roll out updates to its Gemini Audio suite while the promised Gemini 3.5 Pro model remains unreleased since its June launch window.