SANTA CLARA — In a major move to advance agentic AI, NVIDIA has announced the introduction of its Llama Nemotron family of large language models (LLMs) and Cosmos Nemotron vision language models (VLMs), designed to enhance AI agents and their capabilities. These models, optimized for enterprise use, aim to address the growing need for efficient, reliable AI systems across various sectors, from customer support to fraud detection and autonomous vehicles.

The Llama Nemotron models, built upon the widely used Llama foundation models, offer tailored solutions for AI agent development. These models provide enhanced capabilities for tasks like instruction following, coding, and math while maintaining computational efficiency across a variety of devices, from PCs to data centers. The models will be available through NVIDIA’s microservices and API catalog, providing scalable solutions for enterprises to develop their own AI agents.

The collaboration between NVIDIA and industry leaders like SAP and ServiceNow signals the significant potential of agentic AI to transform enterprise productivity. By utilizing NVIDIA’s advanced models, businesses can build customized AI agents that work together to solve complex tasks, optimizing efficiency across various functions. The models also integrate seamlessly with NVIDIA NeMo, a tool that enables customization for specific business domains, further empowering businesses to innovate and streamline operations.

NVIDIA’s new models come in Nano, Super, and Ultra sizes, catering to different performance and accuracy requirements. These models will be accessible through various platforms, ensuring that enterprises of all sizes can harness their power to drive productivity and solve problems efficiently.

For developers and enterprises, NVIDIA offers the flexibility to customize models for specific needs and integrate them into their workflows with the help of the NVIDIA AI Enterprise software platform. The models are available for download via build.nvidia.com and Hugging Face, with free access for members of the NVIDIA Developer Program.

See notice regarding software product information.