🎨 Microsoft Releases 2 New Models from the MAI Family
The company has introduced MAI-Image-2.5-Pro for high-quality image generation with precise text rendering, and MAI-Voice-2-Flash for ultra-fast speech synthesis in 15 languages. Integration into Microsoft services is already reducing computational costs.
🌍 Microsoft is transitioning to its own MAI architecture in key products, which allows for optimizing inference costs by 84-89% and increasing the accuracy of multimodal agents.
👤 Specialized models allow for choosing the optimal balance between speed and quality when developing custom AI applications.
Source 1: https://ai.azure.com/catalog/models?source=microsoft Source 2: https://ai.azure.com/catalog/models/MAI-Image-2.5-Pro