Nebius Cuts 2026 Revenue Forecast as Data Centre Delays Offset 571% AI Cloud Growth

Amsterdam, November 13, 2026: Nebius Group has lowered its full-year 2026 revenue forecast after delays in bringing additional data centre computing capacity online, even as the artificial intelligence cloud infrastructure provider reported a sixfold increase in quarterly revenue and strong customer demand.

The Nasdaq-listed company now expects 2026 revenue of $3 billion to $3.2 billion, compared with its previous forecast of $3 billion to $3.4 billion.

The revision follows delays affecting third-party data centre capacity that was originally expected to become operational before the end of 2026. Some of that capacity is now scheduled to become available during the first half of 2027.

Nebius reported third-quarter revenue of $980.8 million, an increase of 571% from $146.1 million a year earlier. Revenue exceeded market expectations of approximately $907.5 million.

However, the strong operating performance was overshadowed by concerns about construction timelines, the rising cost of AI infrastructure and the company's substantial investment requirements.

Nebius also increased its 2026 capital expenditure forecast to approximately $27 billion from its previous estimate of $20 billion to $25 billion.

For 2027, the company expects revenue of $12 billion to $13 billion and capital expenditure of $40 billion to $45 billion.

The contrasting outlook highlights a central challenge facing AI infrastructure providers: customer demand can remain strong even when delays in electricity supply, construction and computing capacity prevent companies from recognising revenue as quickly as planned.

Nebius Reduces 2026 Revenue Guidance to $3 Billion–$3.2 Billion

Nebius announced its revised outlook alongside its third-quarter results on November 12.

The company lowered the upper end of its full-year revenue forecast by $200 million.

Its previous guidance anticipated revenue between $3 billion and $3.4 billion.

The revised range is $3 billion to $3.2 billion.

The midpoint has therefore declined from $3.2 billion to $3.1 billion.

Management attributed the adjustment primarily to delays in capacity provided through third-party infrastructure arrangements.

These delays affect the timing of revenue recognition rather than necessarily indicating a reduction in contracted customer demand.

Nebius said the agreements supporting the delayed capacity remain in place.

Nevertheless, the change demonstrates how operational timelines can influence financial results in the capital-intensive AI cloud industry.

Nebius Q3 2026 Financial Results: Key Highlights

Financial Indicator Reported Result or Guidance
Q3 2026 revenue $980.8 million
Year-on-year revenue growth 571%
Q3 AI cloud revenue $973.8 million
AI cloud revenue growth 643%
Q3 adjusted EBITDA $465.8 million
Q3 net loss from continuing operations $37.5 million
Q3 adjusted net income $90.9 million
September annualised revenue run rate $6.4 billion
Remaining performance obligations $38.7 billion
Revised 2026 revenue forecast $3 billion–$3.2 billion
Previous 2026 revenue forecast $3 billion–$3.4 billion
Revised 2026 capital expenditure forecast Approximately $27 billion
2027 revenue forecast $12 billion–$13 billion
2027 capital expenditure forecast $40 billion–$45 billion

The results show exceptionally rapid revenue growth alongside continued investment requirements and delays in infrastructure deployment.

Quarterly Revenue Surges 571% to $980.8 Million

Nebius reported consolidated revenue of $980.8 million for the quarter ended September 30, 2026.

This compared with $146.1 million in the corresponding quarter of 2025.

The increase represented 571% year-on-year growth.

Revenue also increased approximately 68% sequentially from the June quarter.

The expansion was supported by new computing capacity becoming available and revenue generated through partnerships involving third-party infrastructure.

Nebius provides cloud computing services designed for artificial intelligence training, inference and related workloads.

Its customers include AI developers, technology businesses and enterprises requiring high-performance computing capacity.

The results demonstrate strong demand for its services.

However, revenue growth must be evaluated alongside the composition of sales, profitability and capital investment.

AI Cloud Revenue Reaches $973.8 Million

Nebius' AI cloud business generated $973.8 million in third-quarter revenue.

This represented growth of approximately 643% compared with the previous year.

The segment accounted for about 99% of total group revenue.

The figures demonstrate the increasing concentration of Nebius' business around AI cloud infrastructure.

The company supplies computing capacity and related cloud services used to develop and operate artificial intelligence systems.

Its platform supports applications ranging from model training to production inference.

The rapid growth reflects strong customer demand for advanced computing resources.

However, the company's ability to meet that demand depends on expanding operational capacity.

Non-Recurring Revenue Contributes $325 Million

Approximately $325 million of Nebius' third-quarter revenue came from non-recurring arrangements associated with its asset-light business model.

Under these arrangements, partners finance infrastructure while Nebius deploys its technology architecture and cloud platform.

The revenue reflects an initial stage of these deployments.

This distinction is important because non-recurring revenue may not repeat at the same level in subsequent quarters.

Nebius expects its partnership business to expand, potentially creating more regular revenue opportunities over time.

However, the timing of individual arrangements may produce fluctuations in quarterly results.

Investors must therefore distinguish between recurring cloud consumption and revenue associated with initial infrastructure deployments.

Data Centre Delays Push Revenue Into 2027

The main reason for the reduced 2026 outlook is the postponement of third-party computing capacity.

Nebius had expected some of this infrastructure to become operational before the end of the year.

However, delays have pushed the expected availability into the first half of 2027.

The company said the underlying contracts remain valid.

Chief Executive Officer Arkady Volozh indicated that delays are a feature of the rapidly expanding AI infrastructure industry.

He also emphasised that the delayed capacity has not been cancelled.

The distinction matters because the company continues to expect revenue from these projects once the infrastructure becomes operational.

However, delays can increase uncertainty about the timing of future cash generation.

New Jersey Infrastructure Faces Power-Related Challenges

One of the delayed projects involves third-party infrastructure in New Jersey.

Nebius is transitioning the site's power supply toward fuel-cell technology provided by Bloom Energy.

The company indicated that initial power-generation equipment had been installed and that capacity was being expanded.

However, the transition has contributed to delays in making the site fully operational.

Power availability is a major constraint for large AI data centres.

High-performance computing equipment requires substantial and reliable electricity supply.

Developers must coordinate electrical infrastructure, cooling systems and computing hardware before commercial operations can begin.

The New Jersey project illustrates how energy infrastructure can determine the timing of AI computing revenue.

Company-Owned Data Centres Remain on Schedule

Nebius distinguished between delays at third-party facilities and progress at its own sites.

Management said company-owned data centres under development were operating on schedule or ahead of their planned timelines.

The company increasingly prefers developing its own facilities because direct ownership provides greater control over execution and operating economics.

Nebius expects most of its next gigawatt of computing capacity to come from owned sites.

However, company-owned infrastructure requires substantial upfront investment.

The strategy therefore involves a trade-off between operational control and capital intensity.

Investors will monitor whether the company's expanding owned-site portfolio reduces the risk of future deployment delays.

Annualised Revenue Run Rate Reaches $6.4 Billion

Nebius reported an annualised revenue run rate of $6.4 billion at the end of September.

This compared with approximately $3 billion at the end of June.

The increase represented sequential growth of 113%.

The annualised measure was also more than eleven times the level reported a year earlier.

However, annualised revenue run rate is not the same as recognised annual revenue.

It extrapolates revenue from a particular period and can be influenced by temporary fluctuations.

Nebius indicated that it increasingly considers revenue and profitability more useful indicators as the business expands.

The company also postponed its previous target of exceeding $7 billion in annualised revenue run rate from December to February.

The change reflects the revised capacity deployment schedule.

Remaining Performance Obligations Reach $38.7 Billion

Nebius reported remaining performance obligations of approximately $38.7 billion at the end of September.

This compared with $37.5 billion at the end of June.

Remaining performance obligations represent contracted revenue that has not yet been recognised.

The measure provides visibility into future business commitments.

However, it does not indicate when all revenue will be recorded.

Recognition depends on contract terms, service delivery and the availability of computing capacity.

For Nebius, the large contracted revenue base supports management's view that customer demand remains strong.

Nevertheless, converting those commitments into revenue will require substantial infrastructure investment and operational execution.

Net Loss Narrows to $37.5 Million

Despite rapid revenue growth, Nebius reported a net loss from continuing operations of $37.5 million during the third quarter.

This compared with a loss of $119.6 million a year earlier.

The reduction indicates improving financial performance as the business scales.

However, the company remains loss-making under the reported GAAP measure.

AI infrastructure providers face substantial depreciation, operating and financing expenses.

The cost of advanced computing equipment is particularly significant.

Nebius must therefore generate sufficient revenue over the useful life of its infrastructure to recover investment costs and produce sustainable profits.

The narrower quarterly loss is encouraging, but it does not eliminate the financial risks associated with rapid expansion.

Adjusted EBITDA Reaches $465.8 Million

Nebius reported adjusted EBITDA of $465.8 million during the third quarter.

This compared with an adjusted EBITDA loss of $5.2 million a year earlier.

The improvement reflects stronger operating scale and revenue growth.

Adjusted EBITDA excludes interest, taxes, depreciation, amortisation and certain other adjustments.

It can help assess operating performance before some significant expenses.

However, it is not equivalent to net income or free cash flow.

For an infrastructure-intensive company, depreciation and capital expenditure are particularly important.

Investors must therefore evaluate adjusted EBITDA alongside net losses, financing costs and cash requirements.

Capital Expenditure Forecast Raised to $27 Billion

Nebius increased its 2026 capital expenditure forecast to approximately $27 billion.

The previous guidance was $20 billion to $25 billion.

The increase reflects the scale of investment required to expand AI computing capacity.

Capital expenditure includes spending on computing equipment, data centre infrastructure and associated systems.

Large AI deployments require advanced processors, networking equipment, cooling technology and electrical infrastructure.

These investments are generally made before the company receives the full revenue generated by the capacity.

As a result, rapid growth can create significant financing requirements.

Nebius' revised spending forecast has increased investor attention on the company's balance sheet and funding strategy.

2027 Capital Spending Could Reach $45 Billion

Nebius expects capital expenditure between $40 billion and $45 billion during 2027.

The forecast represents a substantial increase from the company's 2026 spending plans.

Management intends to expand computing capacity rapidly to meet contracted demand.

The company expects to bring more than one gigawatt of additional capacity online annually.

However, the investment programme depends on access to financing, equipment and suitable data centre sites.

Delays in any of these areas could affect the timing of revenue generation.

The company's ability to maintain financial flexibility will therefore be an important consideration.

Customer Prepayments Become Central to Financing

Nebius expects customer prepayments and asset-backed financing to cover approximately 75% to 85% of its planned 2027 capital expenditure.

Customer prepayments allow infrastructure providers to receive funds before delivering the full contracted service.

Asset-backed financing uses specified assets or contractual arrangements to support borrowing.

The company believes these mechanisms can reduce its dependence on raising new equity or unsecured capital.

However, financing structures carry their own risks.

The availability and cost of borrowing depend on contract quality, asset values and financial market conditions.

Customer prepayments may also create obligations to deliver services within agreed timelines.

Investors will monitor whether the funding strategy supports expansion without excessive financial strain.

Stronger Contract Terms Improve Investment Economics

Nebius reported improvements in the commercial terms of its recent customer agreements.

The company said prepayments associated with third-quarter contracts covered more than 100% of the related capital expenditure.

This compares with approximately 50% to 60% in the second quarter.

Management also reported that the expected revenue-based payback period for comparable contracts improved to approximately one year and seven months.

The previous estimate was one year and ten months.

The shorter payback period reflects higher contracted pricing.

However, these calculations apply to specified comparable agreements and assumptions.

They should not be interpreted as a guaranteed payback period for every Nebius infrastructure investment.

Nebius Expects Revenue of $12 Billion–$13 Billion in 2027

Despite reducing its near-term forecast, Nebius introduced a substantially higher revenue outlook for 2027.

The company expects annual revenue between $12 billion and $13 billion.

The forecast implies rapid expansion from the revised 2026 revenue range.

Management said the outlook reflects a conservative approach to the pace of infrastructure deployment.

The projection depends on bringing new computing capacity online and converting contracted demand into recognised revenue.

Nebius also expects to reach pre-tax breakeven by the end of 2027.

However, the forecast remains subject to construction, financing and operational risks.

The company's execution during the next several quarters will be critical to achieving these objectives.

Demand for AI Computing Remains Strong

Nebius maintains that customer demand continues to exceed available computing capacity.

Management indicated that some customers are already seeking capacity for 2028.

The company also reported stronger pricing for certain future capacity arrangements.

This suggests that demand remains robust despite concerns about the sustainability of AI infrastructure investment.

However, demand for computing capacity is not the same as realised revenue.

Providers must first secure infrastructure, install equipment and deliver reliable services.

The company's reduced 2026 guidance illustrates the difference between customer interest and operational availability.

For investors, the key issue is whether Nebius can convert demand into profitable, recurring revenue.

AI Cloud Infrastructure Faces Growing Competition

Nebius operates in a market that includes established hyperscale cloud providers and specialised AI infrastructure companies.

Competitors include CoreWeave and major cloud platforms offering accelerated computing services.

Customers evaluate providers based on price, availability, performance and software capabilities.

Nebius is attempting to differentiate itself through an integrated model combining infrastructure development, cloud software and managed AI services.

However, competition may intensify as additional computing capacity becomes available.

Large technology companies also possess substantial financial resources and established customer relationships.

Nebius' ability to compete will depend on its infrastructure economics and service quality.

Nvidia Remains Important to the AI Infrastructure Ecosystem

Nvidia's advanced processors are widely used in AI data centres.

Nebius operates within the broader ecosystem of companies building infrastructure around accelerated computing technology.

Nvidia has also invested in Nebius, strengthening the relationship between the two businesses.

However, reliance on advanced computing hardware creates exposure to product availability, pricing and technology transitions.

AI infrastructure providers must coordinate equipment purchases with data centre construction and power availability.

A delay in one component can affect the entire deployment schedule.

Nebius' expansion therefore depends on the performance of a complex technology supply chain.

Microsoft and Meta Contracts Support Long-Term Demand

Nebius has entered into infrastructure arrangements involving major technology customers, including Microsoft and Meta.

Such agreements can provide visibility into future demand.

They may also support financing arrangements when counterparties have strong credit profiles.

However, contracts must be fulfilled according to their terms before revenue can be recognised.

Infrastructure availability remains essential.

Nebius' funding strategy increasingly relies on customer commitments and financing secured against contracted assets.

The company will need to manage delivery schedules carefully to maintain customer confidence and financial performance.

Data Centre Power Availability Becomes a Major Constraint

Power infrastructure is one of the most important requirements for AI data centres.

Advanced computing systems consume substantial amounts of electricity.

Facilities also need cooling systems, networking equipment and reliable backup arrangements.

Securing sufficient power can involve lengthy planning and construction processes.

Grid connection schedules and local infrastructure limitations may delay projects.

Some developers are exploring alternative power arrangements to accelerate deployment.

However, these solutions can introduce additional engineering and operational complexity.

Nebius' New Jersey delays demonstrate how power-related constraints can affect the timing of commercial operations.

Nebius Shares Fall as Investors Reassess Growth Risks

Nebius shares declined sharply following the November 12 earnings announcement.

The stock fell approximately 16% during trading amid concerns about the reduced revenue outlook and increased capital expenditure.

The decline occurred despite quarterly revenue exceeding analyst expectations.

Investors appeared to focus on the timing of future revenue and the scale of funding required for expansion.

The broader AI infrastructure sector has also faced volatility as markets reassess capital spending and financing risks.

However, short-term share-price movements do not necessarily reflect changes in contracted demand.

Nebius' longer-term valuation will depend on revenue conversion, profitability and capital efficiency.

Outlook for Nebius and the AI Cloud Industry

Nebius enters 2027 with a substantial contracted revenue base and plans for rapid infrastructure expansion.

The company expects annual revenue of $12 billion to $13 billion and capital expenditure of $40 billion to $45 billion.

It also expects customer prepayments and asset-backed financing to cover most of its planned investment.

However, the company must resolve capacity delays and bring new infrastructure into operation.

Power availability, equipment procurement and construction schedules remain important risks.

The broader AI cloud industry faces similar challenges as companies attempt to expand computing capacity at unprecedented scale.

For Nebius, the central objective will be converting strong demand into reliable revenue and sustainable profitability.

Conclusion

Nebius Group has reduced its 2026 revenue forecast to $3 billion–$3.2 billion after delays in third-party data centre capacity pushed some expected operations into 2027.

The revision came despite third-quarter revenue surging 571% to $980.8 million and AI cloud revenue reaching $973.8 million.

The company also reported adjusted EBITDA of $465.8 million and remaining performance obligations of $38.7 billion.

However, Nebius increased its 2026 capital expenditure forecast to approximately $27 billion and expects spending of $40 billion–$45 billion next year.

Management remains confident in customer demand and forecasts revenue of $12 billion–$13 billion for 2027.

The challenge is now execution.

For investors, Nebius' results demonstrate both the enormous commercial opportunity in AI cloud infrastructure and the operational and financial risks associated with building data centres fast enough to meet demand.