Microsoft AI releases real-time transcription and new text-to-speech models for voice agents
MAI-Transcribe-2-Streaming covers 60 languages and returns first results in just over 100 milliseconds.
Microsoft AI has released MAI-Transcribe-2-Streaming, a real-time transcription model that the company says ranks first for accuracy on Artificial Analysis, The Decoder reported on Friday.
The model transcribes 60 languages and delivers its first partial results in just over 100 milliseconds, which Microsoft says lets voice agents respond while a speaker is still mid-sentence. The introductory price through the end of the year is US$0.54 per hour of audio.
Microsoft also released two text-to-speech models. MAI-Voice-2.1 speaks 23 languages in the same voice with a native accent in each, and the MAI-Voice-2.1-Flash variant has 150-millisecond latency and costs US$15 per million characters instead of US$22, according to the report.
Both voice models can clone a voice from a few seconds of reference audio, with built-in safeguards intended to prevent misuse. The models are available through Microsoft Foundry and the MAI Playground.
From ARQS Sponsored
On the practical side, West Coast companies can turn to Northline course catalogue, ARQS security, Live Intelligence and AI receptionist. In the ARQS Store, Small Business Launch Systems, AI-J113 — Prompt Architecture for Business, AI-J114 — Practical AI Workflow Foundations are among the newest tools available.