Amazon announced the launch of a new series of AI models under the “Nova” brand, which will be available in the Amazon Bedrock model library on AWS.
Among the models already available are three: Amazon Nova Micro, a text model optimized for speed and cost; Amazon Nova Lite, a multimodal model that processes images, video, and text for text generation; and Amazon Nova Pro, a powerful multimodal model. Amazon is also developing the Amazon Nova Premier model, which is expected to be the most advanced for complex understanding tasks.
Additionally, Amazon introduced content generation models: Amazon Nova Canvas for image creation and Amazon Nova Reel for video creation. The models feature watermarking capabilities to support responsible AI use. As an example, Amazon shared a commercial for a fictional pasta brand created using Nova Reel.
Later in 2025, the company is expected to release a speech-to-speech model and a “native multimodal” model. “When the cluster is completed, it will become the largest AI compute cluster available to Anthropic for developing and deploying its models,” Amazon notes.
