Nvidia is expanding its open-source AI footprint, releasing the lightweight Nemotron 3.5 Lightning model on August 11, 2026, while reportedly developing a massive 1-trillion-parameter Nemotron 4 model to rival proprietary systems. CEO Jensen Huang has championed open-source AI, arguing it accelerates global innovation and directly drives demand for Nvidia's graphics processing units. This strategy unfolds amid intensifying U.S.-China tech competition and political scrutiny over AI distillation techniques.
Nemotron 3.5 Lightning release
- ▪Nvidia released Nemotron 3.5 Lightning, a lightweight open-source AI model designed to run on a single PC graphics processing unit, on August 11, 2026
- ▪Nvidia released NeMo Switchyard, an open-source model-routing software library designed to automatically direct AI tasks to the cheapest and most suitable models, on August 11, 2026
- ▪Nvidia used a technique called distillation to give Nemotron 3.5 Lightning similar capabilities to its larger Nemotron models
Nemotron 4 development
- ▪Nvidia has not set a release date for Nemotron 4, but employees working on the project stated the model could be ready as early as late fall of 2026
- ▪Nvidia is developing a new AI model family called Nemotron 4, with the largest model expected to have at least 1 trillion parameters, according to employees cited by The Information
Open-source AI chip strategy
- ▪Nvidia CEO Jensen Huang signed an open letter urging the U.S. government to support open models to prevent innovation from moving overseas
- ▪Nvidia CEO Jensen Huang stated that free open-source AI models boost hardware and chip sales because the models must still run on graphics processing units
U.S.-China AI competition
- ▪U.S. politicians questioned whether Chinese AI model developers should face sanctions or restrictions similar to those placed on semiconductor chip sales
- ▪U.S. policymakers raised national security concerns over Kimi K3, an AI model developed by China's Moonshot AI that narrowed the performance gap with leading American models
AI safety consortium formation
- ▪Kari Briski, Nvidia's vice president of generative AI, stated that Nvidia is investing in Nemotron to provide accessible frontier open models that strengthen safety and security
- ▪Nvidia formed an AI safety consortium with Microsoft and other companies in July 2026 to focus on using open models and open-source software for cybersecurity
Story comments
Loading comments…