What happened with NBIS today?
No read yet
Uses news and trading volume to explain the day’s move and how it relates to the market.
Stop guessing. Get the call, a three-month target price, the best and worst cases, and a source for every claim.
Nebius Group’s main business is the Nebius AI cloud, which supplies most of its revenue. AI companies, enterprises and developers use its computing capacity, storage and software to train, adapt and run models. Customers pay as they use services or commit to reserved capacity under fixed contracts. Nebius combines its own hardware and software designs with data-center infrastructure to deliver these services.
Microsoft and Meta have signed significant long-term contracts for AI computing capacity, though these purchases may sit outside Nebius’s primary cloud services. Such contracts support infrastructure investment but require substantial spending before delivery and ongoing service commitments. The group also operates TripleTen, an education business that earns revenue from students, and Avride, an early-stage autonomous-vehicle business with a limited revenue contribution.
The core business depends on customers needing enough AI computing capacity to use the infrastructure Nebius builds. Expansion requires additional financing, NVIDIA GPU supplies, suitable sites and reliable power. Data-center projects need permits and regulatory approvals, while sanctions and semiconductor export controls constrain equipment sourcing and which international customers it can serve.
The risk section is the longest, dullest part of the filing. This keeps only what can really hurt you: where it hits, how hard, when it plays out.
Every company claims an edge. Each one is checked against hard numbers like margins and returns: which hold, and which line gives way first.
Nebius Group’s growth depends on AI compute demand, GPU supply and bringing powered data centers online, with privately held Crusoe and Lambda also important rivals to watch.
CoreWeave competes with Nebius Group for enterprises and AI labs paying for GPU computing and managed software to train and run AI models. Watch CoreWeave’s committed cloud contracts for evidence of how much spending these customers are locking in.
Microsoft supports Nebius Group’s expansion by buying substantial long-term AI compute capacity, while Azure competes to sell AI cloud services to customers. Those capacity contracts require upfront investment and ongoing delivery, so Microsoft’s purchases bring execution obligations as well as revenue opportunities.
Nebius Group’s expansion depends on NVIDIA supplying the GPUs and networking that power the AI computing it sells to customers. Watch availability and new-generation deployments: supply constraints can raise Nebius’s hardware costs or limit its ability to meet customer commitments.
Meta Platforms’ multi-year, multi-billion-dollar AI compute purchases support Nebius Group’s infrastructure growth; Meta earns substantially all its revenue from marketers buying ads on its apps. Delivering that capacity requires substantial upfront investment and long-term support from Nebius, rather than simply signing the contracts.
IREN competes with Nebius Group to sell AI training and inference capacity, combining data centers, computing and software in one platform. Watch how much of IREN’s secured power becomes operating AI cloud capacity, not just the size of its power commitments.
Oracle competes with Nebius Group for customers paying for cloud infrastructure to run AI workloads. Its expansion of AI-specific cloud offerings is the relevant rivalry, rather than its hardware and services businesses.
WhiteFiber, Inc. competes with Nebius Group for AI developers paying for GPU capacity to train and run models. Its GPU cloud services—not its separate business hosting customers’ equipment—are the revenue line closest to Nebius’s offering.
ChronoScale Holdings Corporation competes with Nebius Group to sell dedicated NVIDIA GPU capacity under compute-service contracts. Its Together AI agreement charges per GPU per hour and is billed monthly, making those compute fees the concrete business to watch.
Amazon’s AWS competes with Nebius Group to sell cloud computing and AI-specific infrastructure to customers. Watch AWS rather than Amazon’s retail operations for the business that overlaps with Nebius.
Alphabet’s Google Cloud competes with Nebius Group for customers paying to train and run AI models in the cloud. Google Cloud is the relevant business to follow, distinct from Google Services’ advertising revenue.
DigitalOcean Holdings competes with Nebius Group for technology companies and developers buying GPU capacity, managed container infrastructure through Kubernetes, and AI platform services. Its Gradient AI cloud is worth watching for how it combines customers’ traditional and AI workloads on one platform.
Applied Digital and Nebius Group both depend on large cloud customers’ AI infrastructure spending and access to power, but Applied Digital leases GPU-ready data centers while Nebius sells computing capacity. Watch Applied Digital’s power connections and facility commissioning for the constraints on bringing that infrastructure online.
TeraWulf shares Nebius Group’s exposure to long-term demand for powered AI data centers, earning lease payments from large cloud and enterprise customers while Nebius sells AI computing services. Deliveries at TeraWulf’s Lake Mariner and Abernathy campuses are the concrete capacity milestones to follow.
Core Scientific shares Nebius Group’s exposure to AI infrastructure expansion by developing facilities that house customers’ power-dense computing equipment, while Nebius sells GPU cloud capacity from owned and leased data centers. Watch Core Scientific’s conversion and development of facilities for AI hosting, rather than treating its mining operations as the same business.
Cipher Digital and Nebius Group both need powered facilities to serve large cloud customers: Cipher is developing 600 MW of high-performance computing facilities for tenants, while Nebius expands capacity for AI compute contracts. Watch delivery of Cipher’s two development sites; its revenue through December 31, 2025 came from bitcoin mining, not those facilities.
SharonAI Holdings, Inc. overlaps with Nebius Group in selling on-demand GPU computing and cloud storage to customers training and running AI models, though its cloud platform is concentrated in Australia. Follow that Australian cloud business separately from its U.S. data-center development.
Corvex, Inc. competes with Nebius Group to sell AI customers GPU clusters with networking, storage and managed infrastructure for training and inference. Its H200-class and B200 systems are concrete offerings to follow; its Token Factory was still in development, not an established inference business.
Exascale Labs Holdings Inc. competes with Nebius Group to rent customers dedicated GPU servers and virtual machines for AI training, fine-tuning and inference. Its GPU compute and cluster-management services generate revenue, unlike its complementary data-center and cooling capabilities, which had not generated revenue as of its Annual Report.
Vertiv shares Nebius Group’s exposure to building power-dense AI data centers, selling electrical and cooling systems to infrastructure customers while Nebius sells GPU computing. Vertiv’s power-management and liquid-cooling products track infrastructure categories Nebius needs, but the supplied facts do not establish a direct supply relationship.
Meta Platforms, Inc. earningsMETA
Meta Platforms’ report will show how its ad revenue and infrastructure investment support its long-term AI-compute purchases from Nebius Group.
Read Meta Platforms’ ad revenue alongside its infrastructure investment. Resilient ad revenue with continued investment would show that this largely ad-funded customer has a business base for expanding AI capacity; weaker ad revenue alongside reduced investment would make the demand signal for Nebius Group less reassuring. Infrastructure investment covers many uses and is not a measure of purchases from Nebius Group; the takeaway is the customer’s broader capacity and willingness to invest.
CoreWeave, Inc. earningsCRWV
CoreWeave’s report will show whether enterprises and AI labs are still committing to GPU cloud capacity, the same customers Nebius Group competes for.
Look at CoreWeave’s committed cloud contracts and how they translate into revenue. More commitments with revenue following would show that enterprises and AI labs are willing to lock in GPU spending, supporting the demand environment Nebius Group faces; stalled commitments or weak conversion into revenue would point to friction in customer spending or delivery. The takeaway is whether demand in this AI cloud market is turning into business or remains mainly promised.
Nebius Group N.V. earningsThis company
Nebius Group’s report will show whether AI cloud revenue is keeping pace with the cost of expanding its GPU and data-center capacity.
Read AI cloud revenue alongside adjusted EBITDA, a company measure of operating performance. Revenue growth with improving adjusted EBITDA would show that added computing capacity is bringing in sales and becoming more economical to serve; revenue lagging expansion costs while the measure weakens would show investment running ahead of monetization. Together, these figures indicate whether Nebius Group’s AI cloud expansion is improving its economics or still depends on investment.
Nvidia Corp earningsNVDA
NVIDIA’s report will show how GPU demand and supply could affect Nebius Group’s ability and cost to expand AI cloud capacity.
Read NVIDIA’s data-center business performance alongside its comments on supply and production capacity. Strong demand with improving supply would make the equipment environment more favorable for Nebius Group’s GPU expansion; demand continuing to exceed supply could raise hardware costs or delay equipment, slowing expansion and contract delivery. NVIDIA’s overall supply does not establish Nebius Group’s allocation, so the conclusion is about supply risk, not confirmed deliveries.
