Nebius
Nebius (NASDAQ: NBIS) is a fast-growing AI cloud infrastructure company that provides GPU-accelerated compute, storage, and managed AI services to enterprises, researchers, and AI labs across Europe and North America. …
What Nebius Does
Nebius (NASDAQ: NBIS) is a fast-growing AI cloud infrastructure company that provides GPU-accelerated compute, storage, and managed AI services to enterprises, researchers, and AI labs across Europe and North America. Spun out of Yandex in 2024 and listed on NASDAQ, Nebius reported $399 million in Q1 2026 revenue—a 684% year-over-year increase—and is guiding for $3.0–3.4 billion in full-year 2026 revenue, with an annualised run rate target of $7–9 billion.
NVIDIA invested $2 billion in the company, and Nebius has secured a strategic partnership to deploy more than 5 gigawatts of NVIDIA GPU infrastructure by 2030. Its contracted power capacity exceeds 3.5 GW, with more than 75% owned outright.
Major customers include Microsoft and Meta. The Meta agreement announced on 16 March 2026 is worth up to approximately $27 billion, but the structure matters more than the headline: $12 billion of that is dedicated capacity across multiple locations on one of the first large-scale NVIDIA Vera Rubin deployments, delivered from early 2027, while the remaining $15 billion is capacity Meta has committed to buy across upcoming Nebius clusters only to the extent Nebius has not already sold it to third-party customers.
Headquartered in Amsterdam with data centres in Finland, the Netherlands, and growing U.S. presence, Nebius offers on-demand GPU clusters, managed Kubernetes, object storage, and AI-optimised networking.
The company closed a private offering of convertible senior notes on 23 March 2026 with gross proceeds of approximately $4.34 billion — upsized twice from the $3.75 billion originally announced — to fund its data centre build-out. Nebius positions itself as the European-origin alternative to U.S.-centric hyperscalers, combining sovereign-friendly infrastructure with hyperscale GPU capacity for AI training and inference workloads.
Nebius is the only vertically integrated AI cloud in this category that is both European in origin and publicly listed, which is the whole basis of its pitch against CoreWeave: EU-domiciled capacity and majority-owned power for buyers with data-residency or sovereignty constraints that a US-headquartered provider cannot satisfy on paper. Unlike Crusoe it sells directly to end customers, and unlike Lambda or RunPod it runs a managed stack above the metal rather than renting bare instances.
The buyer caveat is availability, not price. Nebius reported its capacity as effectively sold out in its first-quarter 2026 results and raised 2026 capital expenditure guidance to $20–25 billion from $16–20 billion to build more, so near-term self-serve access is limited and the largest tranches of new capacity are already contracted with deliveries scheduled into 2027.
Read the growth rate above with that in mind — it came off a very small base — though the margin trend is the more durable signal, with adjusted EBITDA margin for the AI business improving to 45% from 24% one quarter earlier. Best fit: European enterprises and research groups that need contracted capacity in EU jurisdictions and can plan procurement quarters ahead rather than buy on demand.
Sign in with your company email to claim and enrich this profile.
How Nebius compares in its category
Read our independently researched buyer's guides to see where Nebius sits against the other leading vendors, how the category works, and what to check before shortlisting.