Microsoft AI launches MAI-Transcribe-2-Streaming and new voice models

Microsoft AI released its first streaming speech-to-text model, MAI-Transcribe-2-Streaming, supporting 60 languages with automatic continuous language detection, alongside two text-to-speech models. The transcription model ranked first for accuracy on both final and partial transcripts in an independent Artificial Analysis evaluation. Introductory pricing and availability in Microsoft Foundry expand low-latency voice agent building blocks.