Meet the Incredible AI Stock That Has Nearly Tripled in 2026 and Could Go Much Higher

Nebius has surged about 175% over the past five months as AI demand drives rapid expansion, with revenue up 684% year over year in Q1 and management targeting a $7 billion to $9 billion annual run rate versus $1.25 billion at the end of 2025. The company now has seven data center sites at 100MW+ power consumption, is backed by Nvidia, and is projected by analysts to reach $11 billion in revenue by the end of 2027. Despite the strong growth story, the stock already trades at 5.3x 2027 sales and remains high risk due to distant profitability.

Analysis

NBIS is becoming a pure-play bottleneck bet on AI capex absorption: the market is no longer pricing whether demand exists, but whether the company can convert power, GPUs, and networking into contracted revenue fast enough to justify its expansion. The second-order winner is clearly NVDA, because neocloud scaling effectively turns Nvidia into the toll collector on every incremental watt deployed; the less obvious beneficiary is the infrastructure ecosystem around high-density data centers, while traditional hyperscalers face a marginal cost of capital problem if external GPU capacity remains cheaper and faster to access.

The key risk is that this is a financing and execution story disguised as a growth story. At a multi-year horizon, the stock is highly sensitive to any sign that utilization lags new capacity, because the equity thesis depends on rapid take-up of freshly built megawatts before depreciation and interest expense compound. If AI model training demand normalizes or customers delay migrations, the market will punish the stock on forward sales multiple compression long before revenue rolls over.

The consensus is likely underestimating how much of NBIS’s move is already a bet on scarcity, not quality. Scarcity trades can extend far beyond fundamentals, but once capacity catch-up starts, leaders in the theme often de-rate even while growing quickly. That means the asymmetry is better expressed with defined risk than through outright common stock after a 175% run: momentum can persist for months, but the first disappointment in deployment cadence or forward bookings could trigger a sharp 20-30% reset.

For META and MSFT, the indirect effect is constructive but modest: if neocloud capacity scales smoothly, it reduces the urgency to build every increment of internal capacity at peak cost, which can ease near-term capex pressure and improve AI experimentation velocity. For INTC, the article’s implied demand for advanced infrastructure is only helpful if it can translate into meaningful share in AI-adjacent hardware; otherwise, this trend reinforces that NVIDIA remains the primary capture point in the stack.

Key Decisions for Investors

• Keep NBIS on a tactical long list only, not a core holding; prefer a 1-3 month momentum window and trim into strength if the stock extends another 15-20% without a new contract or utilization catalyst.
• Buy NVDA on any broad AI-capex pullback as a relative winner from neocloud expansion; the risk/reward is better than NBIS because Nvidia captures the spend without balance-sheet execution risk.
• Structure a pair trade: long NVDA / short NBIS on rallies if NBIS trades above 6x forward sales, betting that infrastructure scarcity will normalize faster than GPU vendor economics.
• Use options rather than stock for NBIS exposure: consider a 3-6 month call spread to express upside while capping downside from any delay in data-center ramp or financing concerns.
• Avoid chasing META/MSFT on this headline alone; if anything, treat any weakness in those names caused by AI capex anxiety as a buying opportunity over 6-12 months.