Microsoft Launches MAI Trilogy: Transcription, Voice Synthesis, Image Generation Breakthroughs

Microsoft launches three MAI models in Foundry: MAI-Transcribe-1 (25-language STT, 2.5x faster), MAI-Voice-1 (custom voice from seconds of audio), MAI-Image-2 (2x faster generation, top-ranked on Aren