Amazon Web Services (AWS) Unveils Nova Family of Multimodal AI Models at re:Invent
Amazon Web Services (AWS), the cloud computing arm of Amazon, has introduced the Nova family of artificial intelligence (AI) models at its ongoing AWS re:Invent conference. The new suite of models includes five large language models (LLMs) with various capabilities, including text, image, and video generation. AWS has also announced that these models come with enhanced intelligence and competitive pricing, making them a powerful addition to the company’s growing AI portfolio.
Overview of the Nova AI Models
AWS unveiled five new AI models under the Nova branding, each designed for different tasks, from generating text to creating images and videos. The models offer enhanced performance, and the new pricing strategy aims to make these tools more accessible for enterprise clients. In addition to the existing models, AWS CEO Andy Jassy revealed that a sixth model, Nova Premier, is scheduled for launch in 2025.
Here’s a breakdown of the five models in the Nova family:
- Nova Micro:
- Text-only generation model.
- Lowest latency in the series.
- Context window of 1,28,000 tokens.
- Nova Lite:
- Multimodal: Accepts text, images, and video as input.
- Text-only output.
- Can be used for a variety of tasks, although it does not generate images or video.
- Nova Pro:
- Most capable multimodal model in the series.
- Accepts text, images, and videos as inputs and can complete a wide range of tasks.
- Context window of 3,00,000 tokens.
In addition to these three LLM models, two creative content generation models were also introduced:
- Nova Canvas:
- Image generation model.
- Accepts both text and image inputs to generate creative visuals.
- Primarily designed for advertising, marketing, and entertainment industries.
- Nova Reel:
- Video generation model.
- Can generate short videos from text and image prompts.
- Users can control camera motion using natural language prompts.
Availability and Access
The Nova AI models are available to enterprise clients through AWS’s Amazon Bedrock platform, which is designed to make these AI capabilities more accessible to businesses. The models are positioned to enhance various tasks, from content creation to advanced multimodal applications, and help organizations harness AI for a range of use cases.
Important Highlights of Nova Family of AI Models
- Five Models for Different Needs: The Nova family includes models focused on text generation, image creation, and video production, catering to diverse enterprise needs.
- Improved Performance: The models are designed to be more intelligent with enhanced performance, enabling more efficient and accurate results.
- Multimodal Capabilities: With models like Nova Lite and Nova Pro, AWS offers advanced multimodal AI that accepts both text, image, and video inputs.
- Creative Tools: Nova Canvas and Nova Reel open up new possibilities for creative content generation, specifically targeting the marketing, advertising, and entertainment industries.
- Availability on Amazon Bedrock: AWS clients can access these models via Amazon Bedrock, which is designed for enterprise-level deployment.
Looking Ahead: Nova Premier
In 2025, AWS plans to release a sixth model, Nova Premier, which is expected to bring even more advanced features and capabilities to the Nova family. This upcoming model is poised to further expand AWS’s AI offerings and strengthen its position in the growing AI space.
