NewsTree24 | November 14, 2026 — AI cloud provider Nebius reported a sharp rise in third-quarter revenue, but the results also highlighted a key constraint facing the fast-growing AI infrastructure market: announced demand does not automatically become usable computing capacity on schedule.
Revenue rose to $980.8 million
In results published on November 12, Nebius Group reported consolidated revenue of $980.8 million for the three months ended September 30, 2026, compared with $146.1 million in the same quarter a year earlier — a 571% increase. For the first nine months of 2026, consolidated revenue was $1.962 billion, up 549% year over year.
The company reported a $37.5 million net loss from continuing operations for the quarter, narrower than the $119.6 million loss a year earlier. It also reported adjusted EBITDA of $465.8 million. Adjusted EBITDA is a non-GAAP measure, so it should be considered alongside the company’s reported operating and net-income figures rather than treated as a substitute for them.
Data-centre capacity is the near-term constraint
Reuters reported on November 12 that Nebius reduced its 2026 revenue outlook to $3 billion–$3.2 billion after delays in third-party data-centre capacity pushed some expected openings into the first half of 2027. Nebius said the contracts associated with that capacity remained in place and that its own facilities were on schedule or ahead of schedule.
The distinction matters for enterprise customers. A cloud provider can have customer commitments and strong demand while still facing delays in power availability, construction, networking or third-party colocation capacity. Those bottlenecks can shift when a customer can actually train or serve models, even if the commercial agreement has already been signed.
Expansion requires substantial capital
Reuters also reported that Nebius raised its 2026 capital-spending outlook to about $27 billion and expects 2027 capital expenditure of $40 billion–$45 billion. The company said customer prepayments and asset-backed financing are expected to cover a large share of next year’s planned investment. These are forward-looking plans and remain dependent on financing, customer demand and the timely delivery of infrastructure.
The company’s results show both the opportunity and the execution risk in the AI cloud market: revenue can scale rapidly when GPU capacity is available, but the business is capital-intensive and depends on power, land, networking, equipment and construction partners.
What enterprise AI buyers should take from the results
- Validate delivery dates: ask whether capacity is already live, under construction or dependent on a third-party site.
- Check regional availability: confirm where workloads and data will be hosted and whether the location meets residency and latency requirements.
- Review service commitments: examine uptime, ramp schedules, reserved capacity, credits and remedies if a site is delayed.
- Plan for portability: keep model artefacts, data pipelines and deployment processes portable enough to reduce the cost of moving workloads.
For more context, see our earlier coverage of AI cloud capacity planning and the infrastructure demand behind industrial AI.
Sources
- Nebius: Third-quarter 2026 financial results, November 12, 2026.
- Nebius SEC Form 6-K exhibit: Q3 2026 results.
- Reuters: Nebius pares back 2026 revenue forecast on capacity delays, November 12, 2026.
Featured image is an illustrative data-centre photograph, not a Nebius facility. Photo: Kevin Ache via Unsplash.