Nebius Group N.V. (NBIS) Stock Analysis and Metrics

Nebius

NASDAQ: NBIS

$178.14 ▲ +6.37 (+3.71%)

At close: Jul 17, 2026 · 3:59 PM UTC

Financial Ratios

Market Cap59.20 Bn

P/E369.77

P/S67.43

Div. Yield0.00

ROIC (Qtr)0.00

Total Debt (Qtr)8.45 Bn

Revenue Growth (1y) (Qtr)683.89

Add ratio to table…

About

Nebius Group N. V. builds full stack AI cloud infrastructure to support the rapid growth of the artificial intelligence industry. The company provides compute storage managed services and critical software that enable customers to train run and deploy AI models and applications efficiently. It owns and operates data centers in Finland and colocation sites in France Iceland and the United States and offers a proprietary AI cloud platform that integrates hardware and software layers for optimized performance.

Nebius Group N. V. generates revenue primarily from its AI cloud infrastructure offerings which include GPU based compute instances storage solutions and managed services. The company also earns income from its Avride autonomous driving technology division its Toloka data labeling and generation business and its TripleTen education technology platform. Customers pay for usage of compute resources access to data sets and enrollment in technical training programs.

The company operates through the following segments: AI infrastructure Avride Toloka and TripleTen.

• The AI infrastructure segment designs builds and operates data centers develops proprietary hardware and provides a full stack AI cloud platform that offers compute storage networking and managed services for AI model training inference and deployment.

• Avride develops autonomous driving technology for self driving vehicles and delivery robots and has formed partnerships with companies such as Uber Grubhub Rakuten and Hyundai to deploy its solutions in logistics and ride hailing markets.

• Toloka provides scalable high quality data for AI model training fine tuning alignment and evaluation serving AI research labs foundation model developers Fortune 500 companies and GenAI startups through a global network of annotators and domain experts.

• TripleTen operates an education technology platform that offers immersive training tracks in software engineering quality assurance business intelligence data science cybersecurity and user experience design to reskill individuals for technology careers primarily in the United States and Latin America.

Nebius Group N. V. positions itself as a specialized provider of full stack AI cloud infrastructure competing with both dedicated AI infrastructure firms and major cloud providers. Its main competitors include CoreWeave Crusoe Lambda Labs Amazon Web Services Google Cloud Microsoft Azure and Oracle Cloud. The company’s competitive advantages stem from its control over the entire value chain from hardware design to software platform its proprietary data center designs that achieve high energy efficiency its in house server and rack development and its partnerships with leading chip manufacturers such as NVIDIA.

The company serves a diverse customer base that includes independent developers AI native technology companies frontier AI labs and enterprise customers. Specific partners and customers include Uber Grubhub Rakuten and Hyundai for Avride services and AI research labs Fortune 500 companies and GenAI startups for Toloka data solutions.

Investment Thesis

▲ Bull case

Nebius has secured over 3.5 gigawatts of contracted power with more than 75% coming from owned capacity which creates a durable cost advantage as the company scales its AI native hyperscaler model Owned sites reduce reliance on third party power contracts and protect margins from volatile energy prices The Pennsylvania site alone will deliver 1.2 gigawatts by 2030 providing a long term anchor for U S expansion and supporting the pipeline of committed capacity for 2027 and beyond This power ownership translates into lower operating expenses per GPU hour and enables the company to offer competitive pricing while maintaining healthy margins
The Meta agreement provides a $27 billion five year capacity contract that includes a $12 billion committed compute layer and a $15 billion discretionary option which can be sold to third party AI cloud customers at potentially higher market prices This structure allows Nebius to finance new clusters through asset backed lending at attractive rates while retaining upside if demand from enterprises remains strong The optionality effectively de‑riskes the capital intensive build out and creates a natural hedge against any softening in hyperscaler demand The company can thus pursue aggressive capex knowing that a substantial portion of the investment is backed by a creditworthy counterparty
Nebius AI revenue reached a $1.9 billion annualized run rate in Q1 2026 up over 50% from the prior quarter driven by 841% year over year growth in the core AI cloud business The adjusted EBITDA margin for Nebius AI expanded to 45% up from 24% in the previous quarter reflecting operating leverage as revenue outpaces cost growth The group margin is guided to around 40% for the full year with expectations of a rebound in Q3 and Q4 after a temporary Q2 dip due to back end weighted capacity deployment This margin progression indicates that the business model is scaling efficiently and that profitability will improve as higher utilization and pricing power take effect
Recent acquisitions of Tavily Eigen AI and Clarifai were made to accelerate the product roadmap and deepen customer engagement rather than to generate standalone software revenue Eigen AI provides industry leading model level inference optimization while Clarifai adds system level enhancements and Tavily brings agentic search capabilities These additions strengthen the Token Factory offering and enable Nebius to serve emerging workloads such as agentic AI and vertical specific AI applications By integrating these teams Nebius can increase customer lifetime value and lock in enterprise clients who value a full stack solution that spans from bare metal compute to sophisticated software tools
The partnership with NVIDIA Exemplar Cloud status on GB300 for training workloads aligns Nebius hardware roadmap with NVIDIA’s future product cycles including Vera Rubin and Vera CPU platforms This gives Nebius early access to next generation silicon and ensures that its infrastructure remains performance leadership The equity investment of $2 billion from NVIDIA further cements the strategic alliance and provides additional financial flexibility The collaboration also extends to joint software development for inference and agentic workloads which can accelerate time to market for new product features and enhance the overall value proposition of the Nebius AI cloud

Nebius has secured over 3.5 gigawatts of contracted power with more than 75% coming from owned capacity which creates a durable cost advantage as the company scales its AI native hyperscaler model Owned sites reduce reliance on third party power contracts and protect margins from volatile energy prices The Pennsylvania site alone will deliver 1.2 gigawatts by 2030 providing a long term anchor for U S expansion and supporting the pipeline of committed capacity for 2027 and beyond This power ownership translates into lower operating expenses per GPU hour and enables the company to offer competitive pricing while maintaining healthy margins
The Meta agreement provides a $27 billion five year capacity contract that includes a $12 billion committed compute layer and a $15 billion discretionary option which can be sold to third party AI cloud customers at potentially higher market prices This structure allows Nebius to finance new clusters through asset backed lending at attractive rates while retaining upside if demand from enterprises remains strong The optionality effectively de‑riskes the capital intensive build out and creates a natural hedge against any softening in hyperscaler demand The company can thus pursue aggressive capex knowing that a substantial portion of the investment is backed by a creditworthy counterparty
Nebius AI revenue reached a $1.9 billion annualized run rate in Q1 2026 up over 50% from the prior quarter driven by 841% year over year growth in the core AI cloud business The adjusted EBITDA margin for Nebius AI expanded to 45% up from 24% in the previous quarter reflecting operating leverage as revenue outpaces cost growth The group margin is guided to around 40% for the full year with expectations of a rebound in Q3 and Q4 after a temporary Q2 dip due to back end weighted capacity deployment This margin progression indicates that the business model is scaling efficiently and that profitability will improve as higher utilization and pricing power take effect
Recent acquisitions of Tavily Eigen AI and Clarifai were made to accelerate the product roadmap and deepen customer engagement rather than to generate standalone software revenue Eigen AI provides industry leading model level inference optimization while Clarifai adds system level enhancements and Tavily brings agentic search capabilities These additions strengthen the Token Factory offering and enable Nebius to serve emerging workloads such as agentic AI and vertical specific AI applications By integrating these teams Nebius can increase customer lifetime value and lock in enterprise clients who value a full stack solution that spans from bare metal compute to sophisticated software tools
The partnership with NVIDIA Exemplar Cloud status on GB300 for training workloads aligns Nebius hardware roadmap with NVIDIA’s future product cycles including Vera Rubin and Vera CPU platforms This gives Nebius early access to next generation silicon and ensures that its infrastructure remains performance leadership The equity investment of $2 billion from NVIDIA further cements the strategic alliance and provides additional financial flexibility The collaboration also extends to joint software development for inference and agentic workloads which can accelerate time to market for new product features and enhance the overall value proposition of the Nebius AI cloud

▼ Bear case

The aggressive increase in capex guidance to $20 billion to $25 billion raises execution risk as the company must deliver massive infrastructure projects on time and on budget Any delays in power procurement construction or supply chain could push revenue recognition into later periods and weigh on near term margins The company’s own guidance acknowledges a nonlinear margin trajectory with Q2 expected to be lower due to back end weighted capacity deployment suggesting that the benefits of investment will lag the spending If the market softens during this lag period the company could face underutilized assets and pressure on profitability
While the Meta contract provides optionality it also creates a dependence on a single counterparty for a significant portion of future financing capacity If Meta’s own AI spending plans were to change due to macroeconomic headwinds or strategic shifts the $15 billion discretionary layer might not be fully utilized reducing the upside assumed in the bullish case The company would then need to find alternative customers for that capacity at potentially lower prices which could compress margins and increase the cost of financing the build out
Component cost inflation although described as low single digit could accelerate if global semiconductor demand remains tight or if geopolitical tensions disrupt supply chains The company mitigated some of this impact by early procurement in 2025 but future capex tranches may not enjoy the same hedge Persistent cost pressure would directly raise the expense of adding new GPUs and related infrastructure eroding the expected returns on investment and potentially forcing the company to raise prices which could dampen demand
The rapid expansion relies heavily on the assumption that enterprise AI spending will continue to grow at a strong pace A broad based economic slowdown or a shift in corporate priorities away from AI projects could reduce demand from the fintech manufacturing and life science verticals that Nebius highlights as growth drivers In such a scenario the company’s pipeline might not convert to revenue at the historical win rates and the reliance on prepayments could reverse if customers defer or cancel commitments
The acquisitions of Tavily Eigen AI and Clarifai while strategically sound add integration risk Combining teams with different cultures technologies and go to market models could distract from core execution and delay the realization of product synergies If the acquired talent does not stay or if the expected product enhancements fail to materialize the company may have spent capital without generating the anticipated uplift in customer stickiness or pricing power

The aggressive increase in capex guidance to $20 billion to $25 billion raises execution risk as the company must deliver massive infrastructure projects on time and on budget Any delays in power procurement construction or supply chain could push revenue recognition into later periods and weigh on near term margins The company’s own guidance acknowledges a nonlinear margin trajectory with Q2 expected to be lower due to back end weighted capacity deployment suggesting that the benefits of investment will lag the spending If the market softens during this lag period the company could face underutilized assets and pressure on profitability
While the Meta contract provides optionality it also creates a dependence on a single counterparty for a significant portion of future financing capacity If Meta’s own AI spending plans were to change due to macroeconomic headwinds or strategic shifts the $15 billion discretionary layer might not be fully utilized reducing the upside assumed in the bullish case The company would then need to find alternative customers for that capacity at potentially lower prices which could compress margins and increase the cost of financing the build out
Component cost inflation although described as low single digit could accelerate if global semiconductor demand remains tight or if geopolitical tensions disrupt supply chains The company mitigated some of this impact by early procurement in 2025 but future capex tranches may not enjoy the same hedge Persistent cost pressure would directly raise the expense of adding new GPUs and related infrastructure eroding the expected returns on investment and potentially forcing the company to raise prices which could dampen demand
The rapid expansion relies heavily on the assumption that enterprise AI spending will continue to grow at a strong pace A broad based economic slowdown or a shift in corporate priorities away from AI projects could reduce demand from the fintech manufacturing and life science verticals that Nebius highlights as growth drivers In such a scenario the company’s pipeline might not convert to revenue at the historical win rates and the reliance on prepayments could reverse if customers defer or cancel commitments
The acquisitions of Tavily Eigen AI and Clarifai while strategically sound add integration risk Combining teams with different cultures technologies and go to market models could distract from core execution and delay the realization of product synergies If the acquired talent does not stay or if the expected product enhancements fail to materialize the company may have spent capital without generating the anticipated uplift in customer stickiness or pricing power

Peer Comparison

Companies in the Internet Content & Information

S.No.	Ticker	Company	Market Cap	P/E	P/S	Total Debt (Qtr)
1	GOOG	Alphabet Inc.	4,330.11 Bn	27.03	10.25	77.50 Bn
2	META	Meta Platforms, Inc.	1,553.11 Bn	22.00	7.23	58.75 Bn
3	BIDU	Baidu, Inc.	320.91 Bn	2,283.88	22.76	8.95 Bn
4	AGGI	BILI Social International, Inc.	84.82 Bn	-675,355.91	157,792.74	-
5	JOYY	JOYY Inc.	70.39 Bn	33.64	33.13	0.01 Bn
6	NBIS	Nebius Group N.V.	59.20 Bn	369.77	67.43	8.45 Bn
7	RDDT	Reddit, Inc.	37.81 Bn	53.44	15.29	-
8	SJ	Scienjoy Holding Corp	37.35 Bn	-357.67	217.37	-

Research Tools

Developers

About

Investment Thesis

Peer Comparison

Research Tools

Developers

About

Investment Thesis

Peer Comparison

Create an account

Sign In

Reset Password