Nebius Group reported third-quarter 2026 results before the open this morning, dropping numbers on the same week CoreWeave’s beat-and-raise sent that stock up 16% and lifted NBIS 13% in sympathy. The sector tailwind is real. But Nebius has a more specific problem to solve than its neocloud peer: a full-year revenue guide that requires a sharp Q4 acceleration, funded by a financing structure that the market has been skeptical of for months.
Analyst Targets
- Truist Securities (Arvind Ramnani): Buy, $355
- Goldman Sachs (Alexander Duval): Buy, $328
- Bank of America (Tal Liani): Buy, $310
- Rosenblatt (John McPeake): Buy, $304
- Citi (Tyler Radke): Buy, $265 (cut from $324 ahead of results)
- Piper Sandler (James Fish): Neutral, $185 (cut from $237)
- Rothschild & Co Redburn (Alexander Haissl): Sell, $84
The Numbers That Matter
Consensus entering this morning sat at roughly $907 million in Q3 revenue and a loss of $0.63 per share. That consensus implies about 56% sequential growth from Q2’s $582.3 million. Revenue against the $907 million bar, and Q3 exit ARR, are what investors are actually trading.
The math behind the full-year guide is tight. First-half revenue totaled $981.3 million. If Q3 matches consensus, the nine-month total reaches roughly $1.9 billion, leaving the company needing $1.1 to $1.5 billion in Q4 to hit the $3 to $3.4 billion annual target. That Q4 range requires a step-change from the Q3 run rate, and it only happens if capacity brought online late in the year starts billing at scale before December 31.
ARR Is the Real Scoreboard
Nebius defines annualized run-rate revenue as AI cloud revenue in the final month of the quarter, multiplied by twelve. The Q2 figure was $3.0 billion, up from $1.92 billion at March’s end. Truist entered today expecting ARR above $4.8 billion on 730 megawatts of connected power. Management’s own year-end ARR target is $7 to $9 billion, with 800MW to 1GW of connected power by December 31.
Every $1 billion of year-end ARR above guidance is roughly $83 million of additional December revenue, which also counts toward 2026. So the ARR number and the full-year revenue guide are the same variable written two ways. A strong Q3 ARR exit rate is also a forward revenue statement.
The Financing Plan Under the Microscope
This is where Q3 differs structurally from CoreWeave’s report. CoreWeave’s beat answered whether its revenue model is working. Nebius’s report must answer whether its capital model is working. The company has discussed $20 to $25 billion in 2026 capex against roughly $3 billion in full-year revenue, a gap that only closes through customer prepayments, asset-backed debt, and equity. Management has guided to more than $9 billion in customer prepayments for 2026, covering roughly 50 to 60 percent of associated capex under Q2 contract terms.
The credit market’s appetite for AI infrastructure debt is not uniform. Lenders have drawn a sharp line between borrowers with independent revenue and those whose cash flows depend entirely on AI data center demand.
Three weeks before this morning’s call, Nebius secured a $3.5 billion senior secured facility from Citi, priced at Term SOFR plus 2.75 percent and stepping up over time. While the facility is structured as a short-dated bridge, the company disclosed in its SEC filing that the maturity date is October 16, 2027. Management has pointed to more than $40 billion of long-term contracts as a base for asset-backed financing. The question on the 8:00 AM call is how much of the incremental $20 to $25 billion capex program is now fully funded versus contingent on future raises.
Macro and Industry Context
CoreWeave’s Q3 supplied the strongest recent data point for the neocloud sector: revenue of $3.661 billion, up 168% year over year, margins at 10% at the adjusted operating line, and a revenue backlog of about $132.5 billion. That result directly supports Nebius’s demand thesis, since both companies are selling GPU compute to the same universe of AI builders. What CoreWeave’s quarter cannot answer is whether Nebius’s specific financing sequencing holds. That is Nebius’s story to tell.
Forward Scenarios
Bull
Q3 revenue clears $930 million, ARR exits above $5 billion, and management quantifies how much of the remaining 2026 capex is covered by committed sources rather than to-be-raised capital. The $3 to $3.4 billion full-year guide looks achievable, and analysts who cut targets ahead of the report reverse course.
Base
Revenue lands at or just above the $907 million consensus, ARR comes in near Truist’s $4.8 billion expectation, and the financing update confirms the Citi facility plus prepayments cover near-term obligations with longer-term raises flagged but not yet closed. The guide survives. The stock reacts to the ARR trajectory rather than the headline figure.
Bear
Revenue misses, ARR exits below $4.5 billion, and the financing call reveals the Citi facility is a bridge to a dilutive equity raise or higher-cost debt. The $3.4 billion top-end guide becomes implausible, and bears cite the capex-to-revenue gap as a structural rather than transitional problem.
Private credit is reshaping how AI infrastructure gets built across the sector. Broadcom has been arranging more than $50 billion in private financing for custom AI chips, with Apollo and Blackstone among the lenders.
Technical Overlay
NBIS entered this morning’s report trading around $134, having fallen roughly 47% from its June high of approximately $299. The options market was pricing an implied move of about 17% in either direction around today’s report. The stock’s last four post-earnings reactions were -14%, +11%, +23%, and +32%, meaning the distribution of outcomes here is wide. The gap from Friday’s close to the June peak represents the full distance the market still needs to price in if the ARR trajectory proves credible.
Bottom Line
Passing the $907 million revenue bar is necessary but not sufficient. Nebius is being judged on whether its capex plan is financed with enough certainty to make the ARR targets credible, not just aspirational. The Citi facility bought time. What investors need this morning is evidence that the sequenced funding strategy the company describes is converting into committed capital at the pace the buildout requires. That answer comes from the call, not the press release.
