Amazon Unveils Nova: Advanced Multimodal AI Models for Text, Images, and Videos

Amazon Web Services (AWS) has unveiled Nova, a new family of multimodal generative AI models, at its re:Invent conference. The Nova lineup includes four text-generating models—Micro, Lite, Pro, and Premier—along with image and video-generating models, Nova Canvas and Nova Reel. These models are designed for various tasks, with Premier positioned as a "teacher" model for creating custom models. AWS emphasizes the models' speed, cost-efficiency, and built-in safety measures, while also planning future expansions, including a speech-to-speech model by 2025.