發生了什麼事
Startup Fortune reports that Nvidia is nearing an agreement to acquire Hugging Face for $12.9 billion. The report attributes the agreement to The Information, which cited a person with knowledge of the deal, and says Reuters, CNBC and Bloomberg News also reported on the discussions. Nvidia and Hugging Face have not announced a completed transaction.
Startup Fortune reports that Nvidia has agreed to acquire Hugging Face for $12.9 billion, citing The Information’s report that relied on a person familiar with the agreement. Startup Fortune says Reuters and CNBC carried the report, while Bloomberg News later described Nvidia as nearing an agreement after serious conversations in recent weeks. The article does not provide a public Nvidia or Hugging Face announcement confirming that a transaction has been completed.
The reported price would make the purchase Nvidia’s largest acquisition, according to Startup Fortune. The article compares it with Nvidia’s 2020 acquisition of Mellanox, which Nvidia announced at a transaction value of $7 billion. The comparison is presented as evidence of the deal’s scale, but the source does not provide a definitive merger agreement, closing date, financing structure or information about whether the reported $12.9 billion figure represents cash, stock or another form of consideration.
Startup Fortune reports that The Information said Hugging Face was generating more than $150 million in annualized revenue, compared with about $100 million two months earlier. The article also cites Hugging Face’s summer 2026 open-models report, saying public model repositories on the Hub increased from 2.43 million in January to 2.96 million in August, datasets rose from 711,000 to 1 million, and Spaces increased from 1 million to 1.44 million. These figures are reported claims, not independently audited results in the source.
The article says the acquisition would follow an earlier Nvidia approach. Startup Fortune reports that the Financial Times said Hugging Face rejected a $500 million Nvidia investment offer in late 2025 at a $7 billion valuation, because the company did not want one dominant investor to have too much influence. Startup Fortune also reports that Hugging Face attracted interest from other potential buyers, including Salesforce. The source does not establish whether those discussions produced competing bids or whether Nvidia and Hugging Face have signed binding documents.
為什麼這很重要
Hugging Face operates a widely used hub for open models, datasets and developer tools. Nvidia ownership could give the chipmaker influence over a platform used by rival hardware and cloud companies, raising questions about neutrality, rankings, integrations, pricing and deployment defaults.
Hugging Face is described by Startup Fortune as an open-source hub where developers publish, test, discuss and reuse models, datasets and demo code. The article says the company’s website reports that more than 50,000 organizations use the platform and identifies Meta, Google, Microsoft, Amazon, Intel and Nvidia as having a presence there. If accurate, that breadth means the proposed acquisition would involve infrastructure used across competing parts of the AI industry, rather than only a conventional software product.
The central public-interest issue is platform neutrality. Startup Fortune argues that a chipmaker controlling a major model repository could affect how developers discover, evaluate, optimize and deploy AI systems. The article specifically raises the possibility that rankings, hardware support, pricing, defaults or featured integrations could gradually favor Nvidia. Those are potential conflicts of interest, not reported changes that have already occurred. The source provides no evidence that Nvidia has altered Hugging Face’s policies or treated rival companies unfairly.
The transaction would fit Nvidia’s broader effort to extend its influence beyond hardware, as described by Startup Fortune. The article points to CUDA’s role in making Nvidia hardware attractive to developers and to Nvidia’s Nemotron open models as examples of software and model-layer expansion. Hugging Face would add a community and distribution layer to that strategy. The practical consequence could be greater integration between Nvidia’s chips, software and model-development ecosystem, although the source does not quantify any expected technical or financial benefits.
Open models matter because organizations can run them on infrastructure they control instead of relying entirely on a closed application programming interface, according to the article’s framing. Startup Fortune cites Hugging Face CEO Clem Delangue’s comments on CBS’s Face the Nation that the company used an Nvidia-modified version of a Chinese open model while defending itself after a cyberattack. The source also connects Delangue’s comments to a July letter backed by Nvidia and other companies supporting open- models. These points show the strategic context, but they do not demonstrate that an acquisition would preserve or reduce openness.
互動機制：它實際上是如何運作的
以互動方式探索這項發展背後的基礎技術。
Which component of an AI application is the machine-learning model itself?
接下來看什麼
The key uncertainties are whether a definitive agreement exists, whether regulators or investors scrutinize the transaction, and how Nvidia would govern Hugging Face if the acquisition closes. Public confirmation, deal terms and commitments to platform neutrality would be material next steps.
The first priority is independent confirmation. Nvidia and Hugging Face have not publicly announced a completed deal in the source, and the reported agreement rests partly on an anonymous person cited by The Information. A signed transaction announcement, regulatory filing, board approval or direct confirmation from either company would clarify whether the report describes a binding agreement, preliminary negotiations or an offer that could still fail.
If the transaction advances, observers should seek concrete governance commitments covering model hosting, search and ranking, access for rival chipmakers, pricing, -provider integrations and data portability. Startup Fortune identifies these as areas where ownership could create incentives for preferential treatment, but it does not report that Nvidia has proposed changes. Public terms would be needed to distinguish safeguards from general assurances.
The financial terms also warrant scrutiny. The source reports $12.9 billion in value and more than $150 million in annualized revenue, implying a very high valuation relative to reported revenue. That comparison may explain Nvidia’s strategic interest, but the article does not provide audited accounts, profitability, growth quality, retention, liabilities or the precise valuation basis. Those unknowns limit what can be concluded about the economics of the proposed purchase.
Finally, the outcome for developers and organizations will depend on operational decisions after any closing. Relevant signals would include whether Hugging Face continues supporting non-Nvidia hardware, whether rival models remain equally discoverable, whether enterprise customers face new access tiers, and whether the platform’s open-source governance changes. None of those effects is confirmed in the source. The deal could fail, be modified, face review or close without immediate product changes.