Nvidia unveiled a new series of artificial intelligence (AI) models on Tuesday during its GPU Technology Conference (GTC) 2025. Named Llama Nemotron, these latest large language models (LLMs) are focused on reasoning and are intended to serve as a foundation for agentic AI workflows. The Santa Clara-based technology leader stated that these models are designed for developers and businesses, enabling the creation of advanced AI agents capable of operating independently or collaboratively to tackle complex tasks. The Llama Nemotron models are now accessible through Nvidia’s platform and Hugging Face.
In a press release, Nvidia provided details about the new AI models. The Llama Nemotron reasoning models are built on Meta’s Llama 3 series, enhanced with post-training improvements from Nvidia. The company emphasized that this new family of AI models exhibits enhanced abilities in multistep mathematics, coding, reasoning, and intricate decision-making.
Nvidia noted that these enhancements have increased the models’ accuracy by up to 20 percent compared to the original models. Additionally, the inference speed has reportedly improved fivefold compared to similar-sized open-source reasoning models. Nvidia asserted that “the models can tackle more complex reasoning tasks, improve decision-making capabilities, and lower operational costs for businesses.” With these advancements, the LLM can effectively support the development and operation of AI agents.
The Llama Nemotron reasoning models come in three parameter sizes: Nano, Super, and Ultra. The Nano model is ideal for on-device and edge tasks that demand high accuracy. The Super variant strikes a balance, providing high accuracy and throughput on a single GPU. Lastly, the Ultra model is designed for multi-GPU servers, delivering superior agentic accuracy.
The reasoning models were fine-tuned on the Nvidia DGX Cloud, utilizing curated synthetic data produced by the Nemotron platform alongside various open models. Nvidia is also providing the tools, datasets, and optimization techniques used in the development of the Llama Nemotron models to the open-source community.
Additionally, Nvidia is collaborating with enterprise partners to deliver these models to developers and businesses. The reasoning models and NIM microservices can be accessed through Microsoft’s Azure AI Foundry and also via Azure AI Agent Services. SAP is implementing these models in its Business AI solutions and its AI copilot, Joule. Other companies leveraging Llama Nemotron models include ServiceNow, Accenture, and Deloitte.
The Llama Nemotron Nano and Super models, along with NIM microservices, are available to businesses and developers through an application programming interface (API) on Nvidia’s platform and its Hugging Face listing. They are offered under the permissive Nvidia Open Model License Agreement, permitting both research and commercial use.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream formerly known as CIO News LinkedIn Account | The Mainstream formerly known as CIO News Facebook | The Mainstream formerly known as CIO News Youtube | The Mainstream formerly known as CIO News Twitter
About us:
The Mainstream formerly known as CIO News is a premier platform dedicated to delivering latest news, updates, and insights from the tech industry. With its strong foundation of intellectual property and thought leadership, the platform is well-positioned to stay ahead of the curve and lead conversations about how technology shapes our world. From its early days as CIO News to its rebranding as The Mainstream on November 28, 2024, it has been expanding its global reach, targeting key markets in the Middle East & Africa, ASEAN, the USA, and the UK. The Mainstream is a vision to put technology at the center of every conversation, inspiring professionals and organizations to embrace the future of tech.
