At the re:Invent conference, Amazon announced a series of generative AI models called Nova.

Amazon has introduced the new generation of Foundation models called Amazon Nova. Available through Amazon Bedrock, these models are designed to reduce cost and latency for generative AI applications and are specifically aimed at enterprise environments.

Different models for different applications

Amazon Nova offers two main categories of models: understanding models and creative content creation models. The comprehension models process text, images or videos to generate text output. Creative content creation models generate images or videos from text and images.

There are currently three variants available within the understanding models:

• Amazon Nova Micro: a low-latency, text-centric model suitable for applications such as summarization, translation, and brainstorming.
• Amazon Nova Lite: a multimodal model that processes text, images and videos and can be used for customer interactions and document analysis.
• Amazon Nova Pro: an advanced multimodal model that supports complex workflows, such as financial documents and large-scale code reviews.

A fourth model, Amazon Nova Premier, is expected in early 2025 and is intended to support complex thinking tasks.

Creative models include:

• Amazon Nova Canvas: focuses on image creation with advanced editing features such as inpainting and background removal.
• Amazon Nova Reel: a video model for creating short videos from text and image inputs.

Customization for specific industries

Amazon Nova models offer extensive customization options. Companies can refine models to understand industry-specific terminology, ensure brand identity, and deliver optimal performance for their use cases. For example, a law firm can customize models to interpret legal terminology and document structures.

With this launch, Amazon is responding to the growing demand for scalable and flexible AI solutions for businesses.