Nebius has announced a multi-billion dollar agreement to provide dedicated artificial intelligence infrastructure to Microsoft, highlighting the accelerating global competition for the computing power needed to train and deploy advanced AI systems. Under the multi-year arrangement, Nebius will supply capacity from a new data centre in Vineland, New Jersey, with operations scheduled to begin later this year.
The contract marks a significant step for the Amsterdam-headquartered company, which specialises in cloud services built specifically for intensive AI workloads. Nebius designs both hardware and software in-house, offering compute, storage and managed services tailored to the demands of large-scale machine learning. By securing a long-term commitment from one of the world’s largest technology firms, it positions itself as a key player in the infrastructure underpinning the next wave of AI development.
Expanding capacity for AI workloads
AI models are growing in size and complexity, driving unprecedented demand for high-performance data centres. Nebius’s new facility in New Jersey is designed to meet these requirements, providing dedicated computing power for Microsoft’s own AI initiatives and services. According to the company, the economics of the deal will not only support immediate growth but also accelerate its AI cloud business from 2026 onwards.
To finance the necessary capital expenditure, Nebius plans to use cash flow generated by the agreement and to issue debt secured against the contract, taking advantage of Microsoft’s strong credit profile. The company is also considering additional funding options to support faster expansion than initially planned, a signal of the market’s urgency for scalable AI infrastructure.
Strategic move in a competitive market
Nebius has built a global footprint with research and development hubs across Europe, North America and Israel, and maintains stakes in technology ventures such as ClickHouse and Toloka. Its core business is centred on an AI-native cloud platform, offering a full stack of proprietary software and hardware tuned for machine learning tasks ranging from model training to real-time inference.
The agreement with Microsoft reflects a broader trend as technology companies race to secure reliable access to specialised computing resources. As generative AI applications and large language models become more widely adopted, demand for dedicated infrastructure continues to strain global supply chains and drive strategic partnerships between cloud providers and hardware specialists.
Nebius founder and chief executive Arkady Volozh described the contract as the first in a series of long-term deals with major AI laboratories and technology firms. While specific financial details remain confidential, the scale of the commitment underscores how critical high-capacity, purpose-built data centres have become for the deployment of frontier AI systems.
The partnership demonstrates how investment in specialised infrastructure is becoming central to the progress of artificial intelligence itself. With AI workloads growing faster than traditional cloud capacity can accommodate, alliances like the one between Nebius and Microsoft are set to shape the pace and geography of AI innovation in the years ahead.
