Amazon presents Nova, an ecosystem of AI models capable of understanding and generating text, images and video

Dec. 4 (Portaltic/EP) –

Amazon has presented the new generation of artificial intelligence (AI) models Nova, made up of five models – three for processing or understanding and two for generating creative content, such as images, text and video -, which are already available in Amazon Bedrock.

The company has announced these developments within the framework of your annual AWS re:Invent eventwhich began last Monday, in which the technology company’s latest work in generative AI ‘machine learning’ and cloud environments is announced.

Among the most notable features of its presentation is the ecosystem of Nova foundational models (FM), which have the ability to process text, images and videos, as well as understand graphics and other multimedia content.

Amazon Nova meets five models, three for understanding and two for content generation, and all of them are now available on the platform for developers of generative AI applications with basic models Amazon Bedrock, as has advanced the technology firm.

Amazon Nova Micro is a text-only model that offers lower latency responses at a low cost, while Amazon Nova Lite It has greater agility to process image, video and text inputs.

On the other hand, Amazon Nova Pro has been presented as a highly capable multimodal model that combines precision and speed for a wide range of tasks. Nova Canvas and Nova Reel They are state-of-the-art image and video generation models, respectively.

Finally, Amazon has recognized that Nova Premier It is the “most capable” multimodal model of all and is intended for complex reasoning tasks. It can also be used as a guide for the development of custom models, a feature that will arrive in the first quarter of 2025.

All models support a wide range of tasks in 200 languages -although they are optimized for English, German, Spanish, French, Italian, Korean, Arabic, Chinese, Russian, Hindi, Dutch and Turkish, among others- and multiple modalities. They can also be adjusted based on a range of preferences, so they “learn what customers care about most from their own data (including text, images, and videos), while Amazon Bedrock trains a custom model that will provide personalized responses,” according to the statement.

The technology company has indicated that the new models “are designed to help internal and external developers“, as well as facing challenges in application development, in the words of Amazon’s Senior Vice President of General Artificial Intelligence, Rohit Prasad.

He has also said that they are aimed at providing “compelling intelligence and content generation” and that they offer significant progress in latency, cost-effectiveness and personalization.

Additionally, the Amazon Nova Micro, Amazon Nova Lite, and Amazon Nova Pro are at least 75 percent cheaper than the top-performing models in their respective intelligence classes on Amazon Bedrock.

The brand has also announced that it will present two new Amazon Nova models in 2025, which are a speech-to-speech conversion modelwhich will comprise voice input transmitted in natural language, interpreting verbal and non-verbal signals; and a native multimodal one, which will be able to process text, images, audio and video, both as input and output value.