Friday, September 4, 2026

Beyond GPUs to the AI Model Distribution Network: NVIDIA to Acquire Hugging Face for $13 Billion

Input
2026-09-03 21:21:33
Updated
2026-09-03 21:21:33
Financial News New York — Lee Byung-chul, Correspondent】 NVIDIA is set to acquire artificial intelligence (AI) model platform Hugging Face for $13 billion, or approximately 18 trillion won. Having already dominated the AI semiconductor market, NVIDIA is expanding its influence across the broader AI ecosystem by adding the core distribution network for AI models to its data-center infrastructure portfolio.
NVIDIA agreed on the 3rd (local time) to acquire Hugging Face for $13 billion. The company plans to complete the transaction by 2027, subject to regulatory review.
The deal is the largest corporate acquisition in NVIDIA’s history. It is nearly twice the $6.9 billion NVIDIA paid for network technology company Mellanox Technologies in 2020, previously its largest acquisition. While the Mellanox deal served as a springboard for NVIDIA’s expansion from a semiconductor company into a comprehensive data-center infrastructure provider, the latest transaction broadens its reach into the AI model ecosystem.
Hugging Face is a key platform in the AI industry. It hosts approximately 3 million AI models, 500,000 datasets, and 1 million AI applications. More than 18 million developers and over 200,000 companies worldwide use the platform.
In particular, Hugging Face is a leading platform for “open-weight” AI, which allows users to download and modify models and run them on their own hardware. NVIDIA CEO Jensen Huang said, "Open weights allow startups, enterprises, universities, and public institutions to use advanced AI capabilities without training every model from scratch or paying for the cost of cutting-edge models."
The expansion of the open AI ecosystem could generate new demand for NVIDIA semiconductors. As major NVIDIA customers OpenAI and Anthropic develop their own AI chips, a growing number of startups and companies using open models could help NVIDIA reduce its dependence on a small number of large customers and broaden the demand base for graphics processing units (GPUs).
To support this effort, NVIDIA has pledged hundreds of billions of dollars in investment, lending, and financial guarantees across the AI ecosystem. It has provided funding not only to open-model developer Reflection AI but also to AI infrastructure companies such as CoreWeave and Nebius. The strategy is to expand demand for its GPUs by growing both AI models and the infrastructure that runs them.
However, the acquisition of Hugging Face, which gives NVIDIA control of a key distribution network for AI models, is expected to make regulatory review a significant variable. NVIDIA said it would maintain Hugging Face as “an open platform for the entire AI ecosystem” after the acquisition and allow developers to freely choose their models, cloud services, and semiconductors.
Hugging Face rejected NVIDIA’s $500 million investment proposal late last year, citing concerns about the growing influence of a particular investor. At the time, the company was valued at approximately $7 billion. Just one year later, NVIDIA offered nearly twice that amount to acquire the entire company.
NVIDIA CEO Jensen Huang. Photo=Yonhap News Agency



[email protected] By Lee Byung-chul, Correspondent Reporter