TLDR Tech / Design
Microsoft Launches New High-speed Voice and Image Models
Microsoft launched three AI models — MAI-Image-2, MAI-Transcribe-1, and MAI-Voice-1 — optimized for image generation, speech transcription, and synthetic voice, respectively. MAI-Image-2 produces images up to 1,024×1,024 pixels and runs twice as fast as its predecessor, while MAI-Transcribe-1 transcribes speech 2.5 times faster and achieved a 3.9% word error rate across 25 languages. All three models are available via Microsoft Foundry and are being rolled out across products such as Bing, PowerPoint, and Copilot Audio Expressions.
Read the full story