Nebius lands major infrastructure deals with Meta / Microsoft, cementing its role in the AI cloud backend Nebius’ expansion got a big boost today after sealing lucrative GPU-based infrastructure contracts with Meta and Microsoft — marking a major strategic shift in how AI workloads are hosted and delivered globally

Nebius Group has cemented its position as a pivotal player in the AI cloud backend by landing major infrastructure contracts with Meta and Microsoft, including a $17.4 billion five-year deal with Microsoft announced in September 2025 and a $3 billion agreement with Meta revealed in November. These multi-billion-dollar pacts provide dedicated GPU capacity from Nebius's expanding data centers, starting with its new Vineland, New Jersey facility, enabling hyperscalers to scale AI workloads amid surging demand for Nvidia GPUs and specialized AI software. The deals have propelled Nebius's Nasdaq-listed shares up 248% year-to-date, pushing its market cap beyond $25 billion as it targets $7-9 billion in annualized revenue by end-2026.

The Microsoft contract, potentially expanding to $19.4 billion with additional services, underscores Nebius's appeal to big tech for reliable, high-performance AI infrastructure, with capacity deployment slated for late 2025. Meta's five-year commitment, limited by Nebius's available slots due to overwhelming demand, further validates its ML-centric stack combining Nvidia accelerated computing, custom hardware, storage, and managed services for AI model training and deployment. CEO Arkady Volozh highlighted how these agreements fuel aggressive growth, financed through deal cash flows, debt, and strategic options while prioritizing margins over raw revenue.

Operating data centers across the US, UK, Iceland, Finland, and France, Nebius stands out as Europe's largest neocloud provider, with capacity sold to such operators surging 211% to 414 megawatts in the first nine months of 2025. This positions it against US hyperscalers like Amazon and Google, serving AI startups to enterprises with tools for building and optimizing models. The strategic shift reflects broader industry trends where specialized providers like Nebius handle backend GPU hosting, reducing latency and costs for global AI delivery.

These landmark deals mark Nebius's transformation from Yandex spin-off to AI infrastructure powerhouse, poised to capture more hyperscaler capacity as AI compute demands escalate worldwide