Nebius closes 17.4 billion dollar AI infrastructure agreement with Microsoft

Landmark Agreement to Accelerate Cloud Expansion

Nebius Group, a major force in AI cloud infrastructure, has finalized a transformative agreement with Microsoft Azure, marking one of the biggest partnerships in the AI sector to date. Announced on September 8, 2025, the deal is valued at up to $19.4 billion, positioning Nebius as a preferred provider for Microsoft’s AI data requirements over the next several years[2][3].

Strategic Data Center Development

As part of the multi-year contract, Nebius will construct and operate a new, state-of-the-art data center in Vineland, New Jersey[1]. This facility will deliver substantial, dedicated capacity specifically designed to meet the performance and scale demands of advanced AI workloads.

  • Dedicated infrastructure to power large-scale LLM training and inference
  • Substantial energy-efficient design to support sustainable AI growth
  • First of multiple expected long-term contracts with top tech partners

Financial Structure and Growth Plans

According to Nebius CEO Arkady Volozh, the company will fund its expansion through a combination of generated cash flow and secured debt. Nebius is also actively evaluating additional financing avenues to further accelerate capacity growth, anticipating significant revenue increases as more AI-driven companies turn to its advanced solutions[1].

Industry Implications

This partnership highlights the ongoing race among technology giants to secure reliable, scalable, and sustainable AI infrastructure. With demand for services such as ChatGPT and new enterprise AI models continually surging, investments in robust data center facilities are vital for both performance and innovation.

Looking Forward

The Nebius-Microsoft alliance is set to reshape the landscape for AI infrastructure, with implications for cloud adoption, job creation, and future innovation cycles throughout the tech industry.

Latest AI News

Stay Informed with the Latest news and trends in AI