Nebius Group is currently trading at under 7 times its projected 2026 revenue, a steep discount relative to peers in the AI cloud computing space. Analysts forecast the company’s annualized revenue run rate will jump from $551 million in Q3 2025 to between $7 billion and $9 billion by the end of 2026. That trajectory implies a compounded quarterly growth rate exceeding 120%, yet the market has yet to fully reprice the shares to reflect this expansion.

Although Nebius is on track to report triple-digit year-over-year revenue growth for its fourth quarter and full fiscal 2025, the company remains unprofitable. Management has committed over $1.2 billion in capital expenditures this fiscal year to add GPU clusters and data center capacity. Those investments are expected to drive positive operating leverage in 2027, but will keep free cash flow negative through at least mid-2026.

Nebius plans to be among the first cloud providers to commercialize Nvidia’s next-generation Vera Rubin NVL72 GPU platform. The company has secured agreements to deploy more than 5,000 NVL72 units across its New York, Frankfurt and Singapore data centers by Q3 2026. By offering benchmark-validated infrastructure in multiple regions, Nebius aims to address both regulatory requirements and latency constraints for large enterprise customers, differentiating its service from global hyperscale competitors.