Amazon Unveils Nova, New Family of Multimodal AI Models

Amazon Nova is a new family of multimodal generative AI models introduced by Amazon Web Services (AWS). These models are capable of processing text, image, and video prompts, allowing users to understand and generate multimedia content. The Nova models include Micro, Lite, Pro, and Premier, each designed for different levels of performance and cost-effectiveness. They are integrated with Amazon Bedrock, a fully managed service that simplifies the use of high-performing foundation models through a single API.

Amazon Nova models offer several benefits to customers, including:

• Cost-effectiveness: Nova models are at least 75% less expensive than the best performing models in their respective classes.
• Speed: They are the fastest models in their respective intelligence classes.
• Multimodal capabilities: They support a wide range of tasks across 200 languages and multiple modalities (text, image, video).
• Customization: Customers can fine-tune the models using their own proprietary data to boost accuracy.
• Retrieval Augmented Generation (RAG): Ensures the best accuracy by grounding responses in an organization’s own data.
• Agentic applications: Optimized for interacting with proprietary systems and data through multiple APIs to execute multistep tasks.

The Amazon Nova family includes the following models:

• Amazon Nova Micro: A text-only model that delivers the lowest latency responses at very low cost.
• Amazon Nova Lite: A very low-cost multimodal model that is lightning fast for processing image, video, and text inputs.
• Amazon Nova Pro: A highly capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks.
• Amazon Nova Premier: The most capable multimodal model for complex reasoning tasks, available in early 2025.
• Amazon Nova Canvas: A state-of-the-art image generation model.
• Amazon Nova Reel: A state-of-the-art video generation model that can transform a single image input into a brief video.

Amazon Nova models, particularly Amazon Nova Canvas and Amazon Nova Reel, support creative content generation by providing high-quality output, an intuitive API platform, and advanced customization opportunities. These models help Amazon Ads remove barriers for sellers and advertisers, enabling them to create display and video campaigns for more products, experiment with new strategies like keyword-level creative, and generate engaging content. For example, Amazon Nova Reel can create video advertisements, such as the whimsical 'Pasta City' ad for a fictional boxed pasta brand, showcasing the potential for high-quality, creative content generation.