[Breaking news] Amazon Nova 2 Omni (preview version) has been announced #AWSreInvent

Hello, I'm Kakei (@takaakikakei).

I'd like to introduce you to Amazon Nova 2 Omni, which was announced as a preview version at AWS re:Invent 2025.

Official Information

https://aws.amazon.com/about-aws/whats-new/2025/12/amazon-nova-2-omni-preview/

Model Overview

Amazon Nova 2 Omni is a model characterized by advanced multimodal capabilities. It supports text, images, videos, and audio as inputs, and can generate both text and images as outputs. While many models are limited to text and image inputs with text outputs, Amazon Nova 2 Omni can handle a more diverse range of media types.

Additionally, it supports a context window of 1M (1 million) tokens, enabling the processing of large documents.

Nova 2 Omni is in preview with early access available to all Nova Forge customers, and to authorized customers. Please reach out to your AWS account team for access.

Currently, it's in preview version, with early access available only to authorized customers. We're looking forward to its general availability.

Conclusion

Thank you for reading to the end.

The model mentioned at last year's re:Invent has finally appeared as a preview version. As it's currently only available to early access users, information about the actual user experience is still limited, but it can be said to be a very interesting model from a multimodality perspective. We look forward to its official release and general availability.

See you again!

Additional References

https://www.aboutamazon.com/news/aws/aws-agentic-ai-amazon-bedrock-nova-models