What happened
Nvidia has agreed to buy Hugging Face for $13 billion. Hugging Face is a New York-based platform that hosts 3 million open AI models, about 500,000 datasets, and 1 million AI applications.
Why it matters
Nvidia pledged Hugging Face will remain an open platform, letting more than 18 million developers choose models, cloud providers, and chips. The deal could also ease Nvidia's reliance on a few customers like OpenAI and Anthropic by creating new demand for its chips through open-weight models.
What to watch
The deal’s fate hinges on regulatory review, which Nvidia says it expects to pass, though the outcome remains undecided until approval is granted. Watch for the $500 million.
Summaries like this, in your inbox every morning.
Nvidia's acquisition of Hugging Face for $13 billion marks a major step in its strategy to back open AI models. The company's CEO, Jensen Huang, has already supported groups developing open models, such as Reflection AI, and infrastructure providers like CoreWeave and Nebius. This deal aligns with that push and could reduce Nvidia's dependence on a few large customers like OpenAI and Anthropic, which are also developing their own AI processors.
The purchase also comes amid a US political debate over whether to sanction Chinese developers of open AI systems over alleged technology theft. Nvidia has recently advocated for open models, with Huang signing a letter in July calling for US support. The acquisition may be seen as an effort to strengthen the open ecosystem while navigating regulatory scrutiny, though Nvidia expects the deal to be viewed positively.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Google released Gemini Nano Banana 2.1, cutting API image output to $30 per 1 million tokens from Nano Banana…
Google released EmbeddingGemma 2, its first natively multimodal open embedding model on the Gemma 4 architectu…
Meta, Walmart, Stripe and Sierra Technologies are creating the Personal Agent Protocol, introduced by Sierra c…
Caterpillar and CoreWeave shortened the data-labeling and feedback loop for training autonomous construction m…
Mistral AI opened a public preview of Mistral Large 4, its 1.05 trillion-parameter MoE model nicknamed "Le Cho…
Reflection AI started early access to Beam, its first open-weights model, with 501 billion total and 23 billio…
