Amazon Debuts Nova Models in Bedrock: What It Means for AI

Businesses can use AI to create all manner of communications, from reports to customer-facing videos, and Amazon wants to make that easier.

In an announcement on the Amazon website on Tuesday, Amazon unveiled six new Amazon Nova models. These models are available in Amazon’s Bedrock, the managed service that lets companies build generative AI applications. These Nova models add text, image, and video creation tools.

According to the press release, the models vary from basic to comprehensive, and presumably, the prices will reflect those differences. The six new Nova models include:

• Amazon Nova Micro: A low-cost text-only model.
• Amazon Nova Lite: A low-cost option for processing images, video, and text input.
• Amazon Nova Pro: Also multimodal with better accuracy, speed, and a higher cost.
• Amazon Nova Premier: Highest tier multimodal model that is good for complex tasks and creating custom models. This version is expected to be available in Q1 2025.
• Amazon Nova Canvas: Image generation model.
• Amazon Nova Reel: Video generation model.

“Inside Amazon, we have about 1,000 Gen AI applications in motion, and we’ve had a bird’s-eye view of what application builders are still grappling with,” Rohit Prasad, SVP of Amazon Artificial General Intelligence, said in the announcement. “Our new Amazon Nova models are intended to help with these challenges for internal and external builders, and provide compelling intelligence and content generation while also delivering meaningful progress on latency, cost-effectiveness, customization, information grounding, and agentic capabilities.”

Amazon says the Nova models “are incredibly capable, fast, cost-effective,” which means they could be effective in helping companies design new and better AI applications. They support tasks across 200 languages, and Amazon says they’re “75 percent less expensive” than competitive models.

Two additional Amazon Nova models are also expected in 2025: a speech-to-speech model and an “any-to-any” (or multimodal-to-multimodal) model. The Nova models that are available now, and the ones that are coming, will help companies with image, text, and video processing and creation, which can in turn be used to build customer-facing AI applications.