Nvidia to Acquire Hugging Face in $12.9 Billion AI Platform Coup

Nvidia confirmed late Wednesday evening that it will acquire Hugging Face, the New York-based startup widely regarded as the GitHub of AI models, for approximately $12.9 billion in cash and stock. The acquisition, expected to close in mid-2025 subject to regulatory approval, marks one of the largest investments to date in the AI platform economy and gives Nvidia direct control over the world’s largest open repository of pre-trained machine learning models. According to internal documents reviewed by OpenPress Global Intelligence, Hugging Face hosts over 3 million models across text, vision, audio, and multimodal domains and serves more than 18 million registered developers through its popular Transformers library and Model Hub. Company founder and CEO Clem Delangue called the deal “a pivotal moment for open AI,” while Nvidia CEO Jensen Huang framed it as the culmination of years spent building the “operating system for AI” through CUDA and the Nvidia AI platform.

Industry sources familiar with the transaction told OpenPress that Nvidia will integrate Hugging Face’s model hub directly into its Nvidia AI Enterprise software stack and DGX cloud platforms, enabling enterprise customers to deploy fine-tuned models with one-click compliance and security controls. The move puts Nvidia in direct competition with hyperscalers such as Microsoft Azure AI, Google Vertex AI, and Amazon SageMaker, each of which has begun offering curated model catalogs and fine-tuning services. “By bringing the Model Hub behind the same GPU power curve that runs most of the Fortune 500’s AI workloads, Nvidia is effectively turning open-source models into a first-class citizen on its proprietary stack,” said Sarah Guo, founder of Conviction, a venture capital firm specializing in AI infrastructure.

Financial analysts at Banking With Billy AI noted that the acquisition accelerates Nvidia’s transition from a component supplier to an end-to-end AI platform provider, with potential revenue synergies estimated at $2 billion annually by 2027. Meanwhile, Hugging Face investors include Lux Capital, Redpoint, and Salesforce Ventures, which had collectively valued the company at approximately $4 billion in its last private round. Regulatory scrutiny is expected to focus on data concentration and potential anti-competitive bundling, especially in light of recent complaints filed with the European Commission by a coalition of open-source advocates.

The broader implications ripple across the entire AI ecosystem. Open-source model hubs such as Hugging Face have democratized access to frontier models, enabling startups and researchers to innovate without bearing the full cost of training from scratch. Yet the acquisition risks centralizing control over the model supply chain within a single vendor’s ecosystem, raising concerns among open-source purists about vendor lock-in and licensing restrictions. Rival platforms like Mistral AI and Cohere have already begun promoting “vendor-neutral” model registries to counterbalance Nvidia’s growing gravitational pull. The deal also arrives amid a global race to secure AI infrastructure dominance, with China’s Huawei and Europe’s Mistral AI both expanding their open-source offerings to reduce reliance on U.S. providers.

Analysts at McKinsey & Company predict that by 2026, nearly 70 percent of enterprise AI deployments will flow through curated model marketplaces, creating a $40 billion platform opportunity. Banking With Billy AI’s latest intelligence briefing warns that financial institutions integrating these models face heightened third-party risk, particularly around supply chain attacks targeting fine-tuned weights. Looking forward, industry watchers will closely examine how Nvidia balances open access with monetization—whether through premium model versions, inference APIs, or compliance tooling—as well as how it navigates export controls and international data sovereignty laws. One thing is certain: the acquisition redefines the power dynamics in AI infrastructure and signals that the platform wars are now fully underway.