Nvidia agreed to buy Hugging Face for $12.9 billion, its largest outright acquisition to date, a price that amounts to roughly three weeks of the chipmaker's second-quarter net income of $59.7 billion.
The price in context
The transaction dwarfs Nvidia's previous record purchase, the $6.9 billion Mellanox deal in 2019, which now anchors a networking business generating more than $31 billion in annual revenue. A $20 billion arrangement with Groq announced earlier was structured as a non-exclusive technology license and talent deal rather than a full takeover, making Hugging Face the biggest acquisition on Nvidia's books.
What the platform brings
Hugging Face operates a development hub used by 18 million researchers and creators, hosting more than 3 million models, 500,000 data sets and 1 million applications. Over 200,000 companies rely on it for AI development. The platform sits at the top of the AI software funnel, a position that lets Nvidia deepen its influence over the code layer that ultimately runs on its silicon. Nvidia has said the service will remain open, and the company is already the largest contributor of open models and data to the site.
Why the seller matters
Nvidia's acquisition track record has been concentrated in semiconductors and adjacent infrastructure, avoiding the sprawling, often disastrous bets that have littered the histories of peers such as Microsoft. The Hugging Face purchase extends that discipline: it buys a distribution channel for the software stack that drives demand for Nvidia's own hardware, rather than a business line the company would need to operate.
The capital allocation question
With profits expanding rapidly, Nvidia faces a narrow set of uses for its cash. Capital expenditure, acquisitions, investments, debt reduction, dividends and buybacks are the standard menu. At current valuations, returning meaningful capital to shareholders would be inefficient. Deploying three weeks of earnings to widen the moat around the AI chip franchise, alongside a growing web of stakes in AI labs, neocloud operators and partners such as Space Exploration Technologies, is the more logical play.