Nebius wins $17.4 billion Microsoft GPU infrastructure deal
Nebius Group announced on Monday that it has signed a landmark agreement to provide Microsoft with GPU infrastructure capacity in a deal valued at $17.4 billion over five years. The agreement, which could expand to $19.4 billion if Microsoft opts for additional capacity, underscores the booming demand for high-performance computing resources as tech giants accelerate their investments in artificial intelligence.
Following the announcement, Nebius shares soared more than 47% in after-hours trading, reflecting investor confidence in the company’s strengthening position within the AI infrastructure market.
Under the contract, Nebius will supply Microsoft with access to dedicated GPU capacity from its new Vineland, New Jersey data center, which is scheduled to go live later this year. The infrastructure will support Microsoft’s growing need for advanced AI workloads, particularly those requiring Nvidia graphics processing units — the backbone of modern AI model development.
“Nebius is not only securing strong economics through this agreement, but also positioning itself for long-term expansion,” said Arkady Volozh, CEO of Nebius Group. “The deal will help us accelerate the growth of our AI cloud business in 2026 and beyond.”
Nebius specializes in offering AI developers GPU resources, cloud services, storage, and tools to build, fine-tune, and operate machine learning models. Its approach combines cloud-native software architecture with proprietary hardware design, enabling scalable and cost-efficient performance.
The deal also places Nebius in closer competition with CoreWeave, another leading provider of AI infrastructure. Microsoft is currently CoreWeave’s largest customer, although the company earlier this year denied media speculation about contract cancellations.
Amsterdam-based Nebius emerged from the restructuring of Russian technology company Yandex, gaining independence as part of a broader asset split. The agreement with Microsoft marks Nebius’ most significant commercial milestone since its spin-off, signaling its ambition to become a global powerhouse in AI cloud infrastructure.
