AI Business

NVIDIA Invests $26 Billion in Open-Weight AI Models, Pivoting from Chip Maker to AI Developer

NVIDIA's $26 billion investment in open-weight AI model development revealed in SEC filings. The chip giant pivots to become an AI model developer, directly challenging OpenAI and Anthropic.

NVIDIAOpen-WeightAI ModelsInvestmentSEC
※ このページにはアフィリエイトリンクが含まれています。リンク経由でご購入いただくと、運営費の一部として還元されます。

NVIDIA Invests $26 Billion in Open-Weight AI Models, Pivoting from Chip Maker to AI Developer


In March 2026, filings with the U.S. Securities and Exchange Commission (SEC) revealed that NVIDIA plans to invest $26 billion over the next five years in the development of open-weight AI models. This announcement signals a fundamental transformation of the company from a manufacturer and seller of AI chips into a model development company that will directly compete with leading AI labs such as OpenAI, Anthropic, and DeepSeek.


Alongside this investment, NVIDIA released its most powerful open-weight AI model to date, "Nemotron 3 Super." With 128 billion parameters, the model is reported to outperform OpenAI's GPT-OSS on several benchmarks. Bryan Catanzaro, NVIDIA's Vice President of Applied Deep Learning Research, stated, "NVIDIA is serious about open model development and is making great strides."


NVIDIA's strategy aims to further solidify its dominant position across the entire AI ecosystem by providing open-weight models optimized for its own hardware. Open-weight models, where the model weights (parameters) are publicly available, enjoy strong support from startups and research institutions as they can be freely customized and deployed.


However, this strategy also carries significant risks. NVIDIA's major customers — Microsoft, Amazon, and Google — are each investing in their own AI models and developing proprietary chips to reduce their dependence on NVIDIA products. By directly competing with its customers, NVIDIA could accelerate their move away from its products.


The AI industry is currently witnessing an intensifying competition between open-source and closed-source approaches. While Meta leads with its open model Llama, Chinese companies like DeepSeek and Alibaba are rapidly releasing high-performance open models. NVIDIA's full-scale entry is expected to expand the options for American-made open models and significantly impact this competitive landscape.

AI Newsletter

Get the latest AI tools and news delivered weekly

Related Articles