TradingKey - As of late 2025, Nebius has become perhaps the most highly correlated stock with one of the leading sectors within the AI sector: Infrastructure. The most important question today is not so much about whether Nebius will grow, but rather if this growth will support another significant price increase—potentially doubling during 2026.
Nebius’ metamorphosis has been very fast and out of the ordinary. After being spun-off from Yandex due to geopolitical changes, Nebius established itself as an AI infrastructure company and is now focused on providing the highest-performing computing power via GPU-based data centers.
Nebius also now fits directly into one of the strongest secular trends underway globally: the increase in AI-based capital spending. Hyperscalers are building out data center infrastructure and require an enormous amount of compute power, and Nebius has positioned itself nicely as a "picks-and-shovels" provider to these Hyperscaler customers.
Nebius already counts many of the world's largest companies, such as Microsoft and Meta, as its customers with large multi-year contracts that give it strong visibility for ongoing demand.
Recent numbers provide clear evidence of this growth:
Quarterly revenue grew 355% compared to the same quarter last year, total revenue for 2020 should be in the range of $500-550 million, annual recurring revenue at the end of 2026 could be between $7-9 billion.
That would mean that Nebius has the potential to grow multiple times its current revenue level in a short period of time.
Along with this increase in revenue, Nebius has significantly increased their infrastructure growth plans from approximately 1 gigawatt of computing capacity in 2020 to over 2.5 gigawatts of computing capacity planned for 2026 and has almost completely sold out of their near-term capacity.
Therefore this growth is very important:
The demand for this growth is not speculative, it is based on having contracts in place for future growth and on the fact that they are currently not able to meet contracts due to being at or near maximum capacity.
Nebius has developed an interesting business model which is based on a supply/demand imbalance. This "input-stacking" approach allows Nebius to take advantage of the fact that the demand for AI computing exceeds the supply of AI computing resources available today.
Additionally, there is no room for additional data centers as existing facilities are at capacity, so any new DCs will not be available until at least 2023 (we are midway through 2021). Because of this lack of available data center capacity, Nebius has reached an agreement with customers to enter into long-term agreements for access to AI compute resources.
As a result of this dynamic, Nebius is able to grow revenue more based on current execution than on new demand generation (i.e., revenue will continue to grow as quickly as Nebius can develop and build new AI compute and data center capacity).
The argument for Nebius as a target for a doubling in value by 2026 can be traced to several supporting factors:
1) Revenue Growth – The company plans to grow revenue from less than $1 billion to greater than $1 billion in annual recurring revenue — or multi-billion REVENUE — over the next year. This creates a very powerful re-rating mechanism, especially for investors that want to take advantage of growth.
2) Market Position – Nebius is one of a few companies that provide AI-Native Cloud Infrastructure, or “NeoCloud,” therefore Nebius is in an advantageous position to capture from the amount of hyperscale.
3) Investor Sentiment – Stocks (especially those in the technology sector) are generally traded based upon future earnings (momentum) rather than past earnings. Thus, there continues to be room for appreciation of value to occur, as long as future positive investor sentiment continues.
Considering the current trend of positives towards AI Infrastructure, any modest change in overall sentiment will lead to visible value increases.
The compelling growth story does not support the current valuation on the nebius stock.
At 60x times sales today, the current price of nebius is at a very high multiple, and with large net losses of greater than $100 million in the most recent quarterly results (no profitability expected near-term), it is difficult for an investor in this stock to justify such a high price.
The company will continue to need to invest heavily in capital expenditures to grow.
Thus, investors are paying for the potential growth of nebius that they have yet to see come to fruition.
Nebius's execution in 2026 will be ultimately be the biggest determining factor of delivering the aggressive forecasts, not demand or valuation.
Nebius will rely on the execution of:
Even minor delays in building out capacity may result in the timing of revenue being affected to a material extent impacting investor psychology.
Timing is often as important in these high-growth infrastructure-type businesses as demand.
In general, the price trajectory of Nebius may depend less on "fundamentals" and more on overall market risk appetite.Given that the current market is soft on the monetization of AI, investors are requiring that hyperscaler spending leads to sustainable returns.This creates a binary situation:
Nebius is one of the best examples of how the new breed of AI infrastructure trades exhibit characteristics including – money loss illustrious growth rates and increased uncertainty.
The opportunity they offer is clear:
Overall exposure to one of the largest long-term trends impacting our future.
However, they each present similar risk factors:
It is not your opinion that Nebius has a chance to double its stock price by 2026 – there is real demand for their technology, signed contracts with customers, and surprise projections for revenue growth. But this is also a high risk.
At this stage throughout all stages of the AI ecosystem, growth alone is no longer enough for the markets. It requires growth to be credibly generated from established companies with sound operational execution. For Nebius, the contrast between credible; financial, and operational successes versus failures will be the distinction between being a breakout success or an example of the transformation of AI infrastructure.