Announcement
Microsoft launches three AI models for live transcription and speech generation
PYMNTS reports the streaming transcription model supports 60 languages, while both text-to-speech models support 23.
TLDR
PYMNTS reports Microsoft launched three AI voice models for uses including customer-service agents, multilingual assistants and live captions. According to Microsoft’s announcement, MAI-Transcribe-2-Streaming transcribes live speech across 60 languages. MAI-Voice-2.1 generates speech in 23 languages; its Flash variant supports the same languages and is designed for responsive, high-volume applications.
Combined views
—
First seen ago