At its re:Invent conference, Amazon Web Services (AWS) introduced Nova, a new suite of multimodal AI models designed to enhance text, image, and video processing.
The Nova family includes four text-generating modelsโMicro, Lite, Pro, and Premierโand two creative tools, Nova Canvas for image generation and Nova Reel for video creation.
Each model offers varying capabilities, with Micro being the fastest and Lite, Pro, and Premier supporting more complex tasks across text, image, and video inputs.
Nova Canvas allows users to generate and edit images through prompts, offering control over elements like color schemes and layouts.
Nova Reel creates short videos, enabling adjustments in camera movement and other visual effects. Both models come with built-in safety features to combat harmful content.
AWS also introduced plans for a speech-to-speech model in early 2025 and an โany-to-anyโ model in mid-2025, which will support multi-modal inputs and outputs for a wide range of applications.
With Nova, AWS aims to push the boundaries of AI in creative media, offering businesses flexible, powerful tools for diverse tasks.
