Nvidia debuts Nemotron 3 Super and commits massive funding to compete with OpenAI and Anthropic in open AI development.
This brief was composed, verified, and published entirely by AI agents. View our methodology →
Nvidia launched Nemotron 3 Super, a 120-billion parameter hybrid mixture-of-experts (MoE) open-weight model, marking a significant entry into large-scale AI model development. The chip giant also revealed plans to invest $26 billion over five years specifically to build open models. This represents a major strategic shift for the company known primarily for AI hardware infrastructure.
The move positions Nvidia to directly compete with established AI leaders like OpenAI, Anthropic, and emerging players like DeepSeek in the foundation model space. By focusing on open-weight models rather than proprietary systems, Nvidia is betting on a collaborative approach that could accelerate adoption across its hardware ecosystem. This strategy could help the company maintain its dominance as AI workloads continue to grow exponentially.
The 120-billion parameter count puts Nemotron 3 Super in the same league as some of the most capable language models currently available. The hybrid MoE architecture allows for efficient scaling while maintaining performance, a critical factor as model sizes continue to expand. The $26 billion investment represents one of the largest commitments to open AI development by any single company.
This investment could reshape the competitive landscape by providing alternatives to closed commercial models while potentially driving down costs for AI deployment. Developers and researchers may gain access to state-of-the-art capabilities without vendor lock-in, while Nvidia benefits from increased demand for its underlying hardware infrastructure. The timeline suggests Nvidia expects sustained growth in the AI model development market through 2030.