Article Author: Xu Chao
Source: Hard AI
Morgan Stanley released an in-depth report, systematically dissecting the strategic similarities and differences between the two listed Neoclouds: CoreWeave and Nebius.
According to Morgan Stanley analysts, CoreWeave has established industry leadership through its scale advantage, solid execution track record, and high-visibility financing path; while Nebius constructs differentiated option value with richer spot market flexibility, a diversified strategy targeting enterprise clients, and a more flexible capital structure.
The gap in financial data is immediately apparent. In 2025, CoreWeave generated revenue of $5.13 billion (up 168% year-over-year), with EBITDA reaching $3.1 billion and a profit margin of 60%; Nebius recorded revenue of only $530 million (up 479% year-over-year), with an EBITDA loss of $65 million (margin -12%). Yet the growth rates themselves are equally impressive—Morgan Stanley forecasts CoreWeave's revenue to rise to approximately $12.5 billion in 2026, while Nebius is expected to jump to around $3.3 billion, with both companies on a super-high-speed expansion trajectory.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/8da254d3-700b-453b-8b4d-26b321bb27fc.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/8da254d3-700b-453b-8b4d-26b321bb27fc.png"/>
The critical battleground for investors lies in this: Over 98% of CoreWeave's business is bound by multi-year &#34;use-or-pay&#34; long-term contracts, providing strong revenue visibility; however, Nebius's higher exposure to the spot market could offer greater near-term elasticity given the continued strength in GPU pricing. The approximately $550 million in ARR for Nebius outside of Microsoft and Meta contracts comes primarily from on-demand/spot markets and shorter-term contracts. In the current environment of surging demand and constrained supply, these contracts possess significant room for repricing. Robust demand, combined with supply bottlenecks, memory price inflation, and a surge in token consumption driven by autonomous agents (such as OpenClaw), all support a positive trend in recent GPU pricing.
<h2>Spot Market Exposure: Nebius Holds the Advantage, Short-Term Gains from GPU Price Hikes</h2>
One of the most pressing questions for investors is: Who stands to benefit more from rising GPU prices? The answer points to Nebius.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7e24188c-06b0-41d1-a141-d71530025435.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7e24188c-06b0-41d1-a141-d71530025435.png"/>
Nebius's approximately $550 million in ARR (excluding Microsoft and Meta contracts) comes mainly from on-demand/spot markets and customers with short contract terms (most estimated at under one year). These contracts have room for upward repricing against the backdrop of surging demand and tight supply. In contrast, over 98% of CoreWeave's business is locked into multi-year &#34;pay-as-you-go&#34; long-term contracts, securing revenue visibility but sacrificing spot market flexibility.
Morgan Stanley notes that rising memory prices driving capital expenditure inflation, deployment challenges, and increasing compute consumption by autonomous agents like OpenClaw may sustain a strong trend in GPU pricing in the short term. If GPU prices continue to climb, Nebius is likely to outperform.
<h2>Data Center Strategy: Neck-and-Neck, but Structurally Different</h2>
Both CoreWeave and Nebius employ a &#34;hybrid&#34; data center strategy, combining colocation, build-to-suit, and self-owned facilities, though their existing infrastructure structures differ significantly.
CoreWeave currently operates 43 data centers, primarily relying on colocation, but is actively introducing self-built and joint venture models. Examples include a 300MW self-built project in Lancaster, Pennsylvania (initially 100MW, expandable to 300MW) and a 250MW joint venture project in Kenilworth, New Jersey (CoreWeave holds a 15% stake, expected to come online in early 2027).
Nebius already operates a 75MW self-built data center in Mantzala, Finland, and has planned a 300MW project in Birmingham, Alabama (construction starting Q2 2026) and a massive 1,200MW campus in Independence, Missouri (Phase 1 of 250MW planned for completion by October 2027, with full completion expected by end of 2029). Nebius views self-built campuses as the long-term economic optimum as they allow for better cost and operational control.
Both companies treat colocation as a key tool for rapid expansion and expect to increase the proportion of self-built facilities in the future to achieve better economic benefits and operational control.
<h2>Power Capacity: CoreWeave Far Ahead, Nebius Catching Up Fast</h2>
Power is the core resource in the competition for AI compute clouds.
As of the end of 2025, CoreWeave had signed power capacity of 3.1GW, with active power exceeding 850MW; Nebius had signed power capacity of 2GW, with active power at only 170MW—CoreWeave's active power is fully five times that of Nebius.
Looking at medium-to-long-term targets, CoreWeave plans to achieve over 8GW of signed power by 2030, while Nebius aims for 3GW by the end of 2026 and 5GW by 2030.
Morgan Stanley estimates CoreWeave can add approximately 1GW of active power annually, while Nebius will see a net increase of about 600MW annually. This gap largely reflects different historical execution accumulations—CoreWeave has proven its ability to deploy at scale, whereas Nebius's story remains in a relatively early stage.
<h2>Capital Structure and Financing: Clear Path but High Debt for CoreWeave; Flexible but Uncertain for Nebius</h2>
To support massive capital expenditures, both companies require continuous large-scale financing, though their models differ significantly.
CoreWeave relies on Delayed Draw Term Loans (DDTL) as its core financing tool, having developed a mature asset-level debt financing system.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/52dad00c-e09c-42b8-be9a-e40b54471444.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/52dad00c-e09c-42b8-be9a-e40b54471444.png"/>
As of the end of 2025, CoreWeave's total principal debt reached $21.6 billion (of which 45% was DDTL). Interest rates for successive DDTL tranches have gradually declined: DDTL 1.0 (July 2023, $2.3 billion) carried an interest rate of 15%; DDTL 4.0 (March 2026, $8.5 billion) carries a fixed rate of approximately 5.9%. Morgan Stanley expects CoreWeave's debt scale to grow from $21 billion in 2025 to $46 billion in 2026, and further climb to $92 billion by 2028.
Nebius started with a cleaner balance sheet, relying primarily on convertible bonds, equity offerings (ATM program, 25 million shares), and customer prepayment financing. The recently completed $4.338 billion convertible bond issuance (with maturities of $2.588 billion due in 2031 and $1.75 billion due in 2033) has further bolstered liquidity.
Regarding 2026 capital expenditure plans, CoreWeave guides between $30 billion and $35 billion, while Nebius guides between $16 billion and $20 billion. Both have secured $2 billion in strategic investment from NVIDIA.
<h2>Customer Concentration and Market Strategy: Scale First vs. Enterprise Diversification</h2>
CoreWeave's strategic core is &#34;to reach maximum scale as fast as possible.&#34;
Currently, about two-thirds of the company's revenue comes from Microsoft, and it is expected that by the end of 2026, about two-thirds of ARR will come from the combined total of Microsoft, OpenAI, and Meta. This highly concentrated customer structure brings extremely strong revenue visibility but also sows the seeds of concentration risk.
Nebius's strategic direction is &#34;selling AI cloud to a broader range of enterprise clients and AI startups.&#34; Currently, Nebius's backlog is also highly concentrated, but its anchor contracts with Microsoft and Meta are of exceptionally high quality.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/721345e8-1817-4fdc-a2ab-2b57229d75bb.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/721345e8-1817-4fdc-a2ab-2b57229d75bb.png"/>
Specifically, the $15 billion portion in Meta's latest contract belongs to a &#34;remaining capacity floor commitment,&#34; providing Nebius with &#34;ammunition&#34; to sell GPU compute power to SME clients and AI startups with lower credit ratings (CoreWeave has a similar $6 billion floor agreement with NVIDIA).
Nebius aims to achieve $4 billion to $4.5 billion in ARR from Microsoft and Meta contracts by the end of 2026, and gradually move toward a broader client ARR target of $7 billion to $9 billion.
<h2>Software Stack: CoreWeave More Mature, Nebius Pursues Differentiation</h2>
Both companies strive to position themselves as &#34;full-stack AI infrastructure platforms&#34; rather than mere GPU lessors.
CoreWeave's software stack is more mature, particularly in infrastructure orchestration, observability, workload management, and storage. Its Object Storage business ARR has exceeded $100 million (as of Q3 2025), and it pushes software directly to global cloud service providers and enterprise clients through NVIDIA partnerships.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7d96a567-86e2-4af5-86d6-cb57f9089279.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/7d96a567-86e2-4af5-86d6-cb57f9089279.png"/>
Nebius focuses more on building tightly integrated secure AI cloud environments, emphasizing virtualization, tenant isolation, workload orchestration, and inference toolchains, aiming to create a vertically integrated AI cloud platform tailored for enterprise clients.
Overall, CoreWeave leads in software maturity and platform breadth, while Nebius holds long-term potential in differentiated products geared toward the application layer and inference scenarios.
<h2>Profitability: Different Starting Points, Potentially Converging Endpoints</h2>
CoreWeave already possesses strong EBITDA profitability (EBITDA margin of 60.8% in 2025), but due to timing mismatches between capital deployment acceleration and revenue recognition, adjusted EBIT margins face phase-specific compression (expected to drop to 8% in 2026, rebound to 15.9% in 2027, and further rise to 22.4% in 2028).
Nebius is currently still in a significant loss phase (adjusted EBIT margin of -96.8% in 2025), primarily due to inheriting a large employee and operating expense base from Yandex, drag from loss-making assets (such as AVride), and depreciation and interest pressures during massive expansion.
However, Morgan Stanley forecasts Nebius's EBITDA margin to improve rapidly from -12.2% in 2025 to 40.8% in 2026, 62.1% in 2027, and 70.6% in 2028—if Nebius successfully realizes the TCO advantages of self-built data centers and transitions to higher-margin enterprise clients, its long-term profit potential cannot be ignored.
<img src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/94aa321b-8337-4bea-9cb9-cca3871d206e.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://wpimg-wscn.awtmt.com/94aa321b-8337-4bea-9cb9-cca3871d206e.png"/>
<h2>Both Face Valuation Challenges</h2>
Morgan Stanley assigns a &#34;Neutral&#34; rating to both, with target prices below current stock prices, reflecting that current valuations have already priced in optimistic expectations fairly fully.
CoreWeave's core risks lie in its massive debt scale of $46 billion (projected for 2026) and highly concentrated customer structure; Nebius's core risks involve aggressive short-term revenue targets, unproven profitability records, and significant uncertainty in transitioning sales to a broader enterprise client base.
Potential risks common to both companies include normalization of AI demand, internalization of AI compute by major clients, and shortened effective useful life of GPUs. For investors seeking scale and certainty, CoreWeave's execution capability and financial visibility are more attractive; for those pursuing growth elasticity and diversified option value, Nebius's spot market exposure and enterprise cloud strategy offer a differentiated investment logic.

NBIS

CRWV

CLOU

Morgan Stanley's first systematic analysis of the two AI cloud giants reveals that CoreWeave has established industry leadership through scale, a solid execution track record, and high-visibility financing paths; while Nebius builds differentiated option value via richer spot market flexibility, a diversified strategy targeting enterprise clients, and a more flexible capital structure. However, unresolved risks regarding high debt levels and customer concentration loom over both—this battle for AI infrastructure ultimately boils down to a trade-off between certainty and imagination

- Morgan Stanley's report analyzes the strategic differences between two publicly listed companies, CoreWeave and Nebius.  
- CoreWeave leads in scale, securing long-term contracts, while Nebius offers flexibility in the spot market and aims for a diverse client base.  
- Financially, CoreWeave anticipates sales of $5.13 billion in 2025, with a 60% profit margin, whereas Nebius expects $530 million, facing losses, although both are on trajectories for rapid growth.

Wallstreetcn

Nebius

Coreweave

Global X Cloud Computing ETF

Morgan Stanley

Microsoft

Meta Platforms

NVIDIA

OpenAI

MORGAN STANLEY DEP SHS EA REP 1/1000 SHS NON CUM PFD O

MORGAN STANLEY DEP SHS REPSTG 1/1000TH PFD SER Q

MORGAN STANLEY DEP SHS REPS 1/1000 7.125% PFD E

MORGAN STANLEY NON-CUM DEP SH REP 1/1000 PFD SER I

MORGAN STANLEY 4.875 DEP SHS REP 1/1000 NCUM PFD SR L

MORGAN STANLEY 6.5%DEP SHS REPSTG 1/1000 NON CUM SER P

MORGAN STANLEY DEP SHS REPS 1/1000 PFD SER 'A'

MORGAN STANLEY DEP SHS REPR 1/1000TH INT PFD SER 'F' FX

MORGAN STANLEY DEP SHS REP 1/1000 NON-CUM PFD STK SER K

NVIDIA Corporation

CoreWeave vs. Nebius: Which New Force in AI Cloud Reigns Supreme?