Nvidia presents new open source AI models amid growing Chinese supply | Businesses

Nvidia is primarily known for providing chips that companies like OpenAI use to train their proprietary models and charge for them. But it also offers a range of its own models for everything from physics simulations to autonomous vehicles as open source software that can be used by researchers or other companies, with companies like Palantir integrating Nvidia’s model into their products.

Nvidia unveiled the third generation of its major language models this Monday »Nemotron“, intended for writing, programming and other tasks. The smallest of the models, called Nemotron 3 Nano, was launched this Monday, with two other larger versions planned for the first half of 2026.

Nvidia, which has become the world’s most valuable publicly traded company, said the Nemotron 3 Nano is more efficient than its predecessor, meaning it will have a lower operating cost and perform better in long, multi-step tasks.

Nvidia is launching these models at a time when open source offerings from Chinese technology companies such as DeepSeek, Moonshot AI and Alibaba Group Holdings are increasingly used in the technology sector, with companies such as Airbnb touting the use of Alibaba’s Qwen open source model.

At the same time, CNBC and Bloomberg reported that Meta was considering moving to closed-source models, leaving Nvidia as one of the leading open-source solution providers in the United States.

Many US states and government entities have banned the use of Chinese models due to security concerns.

Kari Briski, vice president of enterprise generative AI software at Nvidia, said the company aims to provide a “model that people can trust” and also makes its training data and other tools freely available so government and enterprise users can test them for security and customize them.

“That’s why we treat it like a library,” Briski told Reuters in an interview. “That’s why we’re engaging in this from a software engineering perspective.”