Nebius Soars 202% in 2025 After $22.4B AI Cloud Contracts with Microsoft and Meta
Nebius Group’s stock rallied 202% in 2025 after securing AI cloud infrastructure deals with Microsoft valued up to $19.4B and Meta worth $3B, fully selling out its data center capacity. Management forecasts revenue rising from about $1B to $7-9B in 2026 but cautions that high execution risk could derail growth.
1. Surge Driven by AI Cloud Demand and Data Center Expansion
Nebius Group’s stock has climbed nearly 30% over the past month and 97% over the past six months, reflecting surging demand for its AI-optimized cloud infrastructure. The company has added three new hyperscale data centers in Northern Virginia, Frankfurt and Singapore since July, bringing total global capacity to 1.2 exaflops. Management reports selling out of existing capacity and plans to double server deployments by mid-2026 to meet growing enterprise and research workloads.
2. Strategic Deals with Hyperscalers Boost Revenue Projections
During 2025, Nebius secured two headline agreements: a multi-year contract with a leading cloud provider valued at $19.4 billion and a five-year supply deal with a major social media platform worth approximately $3 billion. These partnerships underpinned a 202% share price rally for the full year and helped company leadership increase revenue guidance from a $1 billion annual run rate at mid-year to an $8 billion target by the end of 2026.
3. Launch of AI Cloud 3.1 with Blackwell Ultra and Global Footprint
In December, Nebius rolled out its AI Cloud 3.1 platform featuring Blackwell Ultra GPU clusters, offering performance improvements of up to 60% over prior generations. The upgrade is supported by a newly activated data center in Seoul and planned deployments in São Paulo and Dubai. Executives anticipate that enhanced performance and geographic reach will accelerate customer onboarding and drive utilization rates above 80% by Q3 2026.
4. Valuation and Execution Risks for Investors
At roughly three times projected 2026 sales, Nebius appears reasonably valued relative to peers in the AI infrastructure sector. However, the company faces high cash burn—over $500 million in the last quarter—and must execute on aggressive expansion plans. Any construction delays, supply-chain bottlenecks for advanced GPUs or a slowdown in enterprise AI spending could pose material headwinds to achieving the lofty $7–9 billion revenue target.