Nvidia unveils Nemotron AI models to enhance AI agents SAP and ServiceNow early adopters

Nvidia has announced new AI model families called Nemotron, designed to advance AI agents through enhanced language and visual processing capabilities.

The big picture: Nvidia’s latest announcement at CES 2025 introduces two key model families – Llama Nemotron for language processing and Cosmos Nemotron for visual understanding – aimed at enabling more sophisticated AI agents that can handle complex enterprise tasks.

• CEO Jensen Huang describes AI agents as “the next robotic industry,” projecting it to become a multibillion-dollar opportunity
• The models are built on the Llama foundation, which has seen over 650 million downloads, indicating strong market validation
• These new models are designed to work across various computing platforms, from edge devices to data centers

Technical capabilities and features: The Nemotron family introduces three distinct model sizes to accommodate different deployment scenarios and performance requirements.

• Nano models focus on real-time applications with low latency, optimized for PCs and edge devices
• Super models deliver high accuracy and exceptional throughput on single GPU systems
• Ultra models target data-center applications requiring maximum performance and accuracy
• All models excel at instruction following, chat functionality, coding, and mathematical operations

Industry partnerships and adoption: Major enterprise software providers are already planning to integrate Nemotron models into their platforms.

• SAP will incorporate the technology into their Joule platform, potentially reaching hundreds of millions of enterprise users
• ServiceNow plans to utilize the models to enhance their AI agent services
• The models will be available through Nvidia NIM microservices, enabling seamless deployment across clouds, data centers, and workstations

Implementation and accessibility: Nvidia has created multiple pathways for enterprises to access and customize the Nemotron models.

• The models will be available as downloadable files and hosted APIs through build.nvidia.com and Hugging Face
• Free access is provided to Nvidia Developer Program members for development and testing
• Enterprise deployment is supported through the Nvidia AI Enterprise software platform
• Customization options include data curation tools and integration with enterprise data through retrieval-augmented generation (RAG) capabilities

Looking ahead: The introduction of Nemotron models represents a significant step toward more capable AI agents, though success will largely depend on enterprise adoption and practical implementation challenges. The technology’s ability to combine language and visual processing could enable more sophisticated automation across industries, but organizations will need to carefully evaluate deployment costs and integration requirements against their specific use cases.