Nebius rockets on AI cloud boom and billionaire backing—but can it outpace the hyperscalers?

Nebius's financial performance in early 2026 has been nothing short of extraordinary, with first-quarter revenue soaring 684% year-over-year to $399 million, driven primarily by its AI cloud segment which alone grew 841% to $390 million. This surge not only exceeded investor expectations but also marked a significant turnaround from a $53.7 million adjusted EBITDA loss to a $129.5 million profit, underscoring the company’s rapid path to profitability. The market responded enthusiastically, with Nebius’s stock price jumping 24% in the days following the earnings announcement, reflecting strong investor confidence in its financial trajectory.

Supporting this explosive growth, Nebius boasts a robust contracted backlog of approximately $46 billion, anchored by massive multi-year deals with Meta ($27 billion) and Microsoft ($17 billion), which provide substantial revenue visibility and a stable foundation for future expansion. The company’s strategy to transition from volatile spot GPU rentals to long-term, multi-year commitments has transformed its revenue streams into predictable and recurring income, as evidenced by a fully booked current fleet and intense competition among four to five customers vying for each new cluster. This shift not only stabilizes cash flow but also justifies Nebius’s ambitious plans to deploy over 5 gigawatts of AI computing capacity by 2030 through multiple gigawatt-scale AI factories across the U.S.

To sustain and accelerate this momentum, Nebius has committed to an aggressive capital expenditure program of $20 billion to $25 billion in fiscal 2026, including the construction of a 300-megawatt data center in Vineland, New Jersey, and a gigawatt-scale AI Factory campus in Missouri. These investments underpin management’s full-year revenue guidance of $3.0 billion to $3.4 billion with an adjusted EBITDA margin target around 40%, despite an expected short-term margin dip in Q2 due to the timing of capacity build-out. Confidence remains high that margins will rebound and improve in subsequent quarters, reflecting a well-calibrated balance between rapid infrastructure scaling and sustained profitability.

Leopold Aschenbrenner's Situational Awareness Fund made a landmark strategic investment by acquiring a $2.6 billion, or 5.6%, stake in Nebius, underscoring a robust vote of confidence in the company’s AI infrastructure leadership. This move not only propelled Nebius’s stock price by 11% in premarket trading and contributed to a 24% rise following strong financial results, but also signaled a broader market momentum fueled by high-profile, tech-savvy investors who understand the critical role of AI infrastructure in the sector’s growth.

The investment by Situational Awareness is part of a larger validation of Nebius’s market position, amplified by Nvidia’s $2 billion commitment through 2030 to co-develop AI inference stacks and GPU monitoring systems. This strategic partnership has helped Nebius’s stock surge over 100% since March, reflecting heightened market confidence driven by cutting-edge collaborations that promise to scale AI capabilities effectively.

Aschenbrenner’s unique background as a former OpenAI researcher turned fund manager epitomizes a new wave of investor sophistication where AI technologists are directly channeling their expertise and capital into foundational AI infrastructure companies like Nebius. This convergence of research insight and investment acumen is reshaping market dynamics, as exemplified by the rapid rise of the $13.7 billion Situational Awareness Fund and its influential stake in Nebius.

Nebius is executing a bold $25 billion capital expenditure strategy to dramatically scale its AI infrastructure, aiming to boost contracted power capacity from just over 2.5GW to an anticipated 4GW by year-end 2026. This rapid expansion includes plans to increase its data center footprint from 7 to 16 facilities within a year, underscoring a commitment to vertical integration where owned capacity now constitutes more than 75% of total contracted power—an essential factor for enhancing long-term margins and operational control.

Strategic acquisitions and facility launches are central to Nebius’s AI inference enhancement, highlighted by the $643 million acquisition of Eigen AI, a MIT-affiliated startup specializing in reducing compute costs for AI inference. Complementing this software-driven approach, Nebius is investing heavily in physical infrastructure, including a 1.2GW AI factory in Independence, Missouri, a 1.2GW site in Pennsylvania, and a 310MW facility in Finland, alongside new colocation sites in Spain and a 300MW data center in New Jersey, collectively expanding its global AI compute footprint.

Beyond raw capacity, Nebius is pioneering AI infrastructure innovation through the launch of AI robotics labs and partnerships with academic institutions like Rowan University, aimed at cultivating a skilled workforce to operate its advanced facilities. The Physical AI Living Lab, developed in collaboration with Nvidia, provides robotics startups access to cutting-edge simulation and cloud resources, signaling Nebius’s strategic pivot from pure hardware expansion to integrating software orchestration and autonomous machine capabilities within its AI ecosystem.

Nebius’s infrastructure build-out is not only vast but also strategically aligned with market demand, as management reports fully booked capacity with multi-year contracts replacing spot purchases. This shift to predictable, long-term revenue streams is supported by a global network of nine new sites under construction across the US and Europe, including Nvidia-powered plants in the UK and a £1.7 billion investment targeting 65 megawatts of AI capacity by 2027, reflecting confidence in sustained customer demand and the company’s ability to deliver scalable AI compute solutions.

Investor sentiment toward Nebius is marked by cautious optimism tempered by the immense competitive pressures from hyperscalers whose capital expenditures dwarf Nebius’s by a factor of eight, as highlighted by both Roman Chernin and Nebius’s co-founder. While the AI infrastructure market is not seen as a bubble, the race to rapidly scale capacity in megawatts, gigawatts, and GPUs is urgent and capital intensive, with limited short-term flexibility—'In the next 6 months, the capital cannot help,' the co-founder admits—fueling mixed confidence among investors who recognize the high stakes involved.

A predominant strategic risk looming over Nebius is the threat of excessive market consolidation, which could concentrate power among a handful of hyperscale giants and marginalize Nebius’s role to merely a physical infrastructure provider for these super companies. Both Roman Chernin and Nebius’s leadership emphasize that while competition is fierce, it is the potential for a consolidated industry—where 'three to five super models, super companies control the world'—that poses the greatest existential challenge, underscoring the delicate balance Nebius must maintain to preserve its growth trajectory.

Despite these risks, there remains a thread of cautious hope rooted in the organic innovation driving market diversification, which Nebius’s co-founder believes could prevent the AI infrastructure landscape from becoming overly consolidated. This optimism is bolstered by significant strategic investments like Leopold Aschenbrenner’s 5.3% stake, which the company views as a vote of confidence but also a reminder to 'keep yourself like down to the ground' and deliver consistent results amid the pressures of rapid scaling and operational complexities such as supply chain and regulatory hurdles.