Nebius 2026 Revenue Guidance Cut and $40–45 Billion Build Plan, November 2026

Nebius posted a 571% Q3 revenue jump but cut its 2026 outlook on data center delivery delays while projecting a $40–45 billion build; shares fell 9.9% to $115.88 on November 13.

Nebius (NASDAQ: NBIS) closed the November 13 trading session down 9.9% at $115.88, the latest leg of what one report described as a two-day cumulative decline of roughly 25% for the stock. The move did not occur in isolation: reports from November described sector peers including Applied Digital and CoreWeave also falling sharply, while multiple reports indicate the broader neocloud sector has declined between 39% and 50% over the preceding month. One outlet attributed the continuation of losses specifically on November 13 to a personnel development at NVIDIA that rattled confidence across AI-infrastructure names; that said, no single published account fully explains the 9.9% session move, and Nebius's shares had already come under sustained pressure after a third-quarter earnings release that combined a revenue beat with a cut to forward guidance.

The earnings report, filed with the SEC as a Form 6-K on November 12, presented a complicated picture. The headline read well: Q3 revenues reportedly surged 571% year over year, beating analyst estimates on AI cloud growth. The forward signals were more sobering. Reuters and multiple other outlets reported that Nebius lowered its 2026 revenue forecast, citing data center capacity delivery delays; at least one report specifically mentioned a Microsoft data center project among those affected. Simultaneously, the company disclosed a $40–45 billion build plan alongside a 2027 capital expenditure outlook that one publication described as having skyrocketed, and set a breakeven target for late 2027. The combination of a trimmed near-term revenue ceiling and a sharply expanded long-term spending commitment redirected market attention from the revenue achievement itself toward execution and delivery risk.

The capital scale involved helps explain why delivery timelines are receiving such close scrutiny. According to Slicast's own compilation of SEC XBRL filings, Nebius's capital expenditure reached $4.07 billion in FY2025 — a 403.4% increase from $808 million in FY2024 — against revenue of $530 million in the same fiscal year, implying a capex-to-revenue ratio of 767% for FY2025, the highest among the seven neoclouds in Slicast's peer group. The company has been methodical in constructing the financing architecture to support this pace: a reported $3.5 billion facility arranged through Citi was described by management as an early tranche of a "sequenced" financing plan, with larger asset-backed loans expected to follow. In October, the CEO publicly stated that Nebius has sufficient capital to build its 1-gigawatt infrastructure target while also acknowledging that "some of the capital sources are temporarily closed." In Q3, the company reportedly signed $12 billion in contracts, which were described as covering its capex program — a potentially meaningful buffer if delivery delays prove temporary rather than structural.

The competitive backdrop raises questions that extend beyond any single quarterly result. Reports in November documented that Microsoft is declining to renew cloud contracts with Oracle and pivoting toward its proprietary Maia 300 chip, in what one outlet characterized as a systematic reduction of hyperscaler reliance on independent neocloud providers. For a sector whose demand thesis rests in part on overflow workloads from the largest cloud buyers, that strategic shift carries real analytical weight even where its direct impact on Nebius's customer base has not been separately confirmed. Elon Musk's public assertion that more than 200 AI neoclouds face "guppy-level death" and a Gartner estimate that 80% of neoclouds could face consolidation add further structural pressure to the sector narrative. Against that, Nebius offered its own counter-data point: the company raised GPU cloud prices 17–21% effective October 1, citing a supply-demand reversal in its favor. The company has also moved to diversify its infrastructure footprint geographically, with a 50-megawatt agreement at the NEOM Oxagon campus in Saudi Arabia through HUMAIN and DataVolt, reported funding for a French data center, and Bloom Energy fuel cell deployment at its 300-megawatt New Jersey facility.

The investment case at this juncture presents a genuine tension between scale and timing. On the opportunity side: a 571% year-over-year revenue surge, $12 billion in reported Q3 contracted deals, rising GPU pricing power, geographic diversification spanning North America, Europe, and the Gulf, and a $3.5 billion credit facility as a financing foundation with larger tranches anticipated. On the risk side: a 2026 guidance cut rooted in capacity delivery delays, a capex intensity of 767% of FY2025 revenue that remains the highest in the neocloud peer group, a sector that has shed roughly 39–50% over the past month, and an unresolved structural question about how aggressively hyperscalers will internalize AI compute capacity. Three signals warrant monitoring in the quarters ahead: first, whether the data center delivery delays resolve on a timeline consistent with the revised 2026 revenue guidance; second, the pace of revenue recovery in early 2027 and its trajectory toward the late-2027 breakeven target; and third, the rate at which Microsoft and other large cloud buyers act on stated intentions to build proprietary AI infrastructure — a decision whose ultimate scope will do much to define how large the addressable market for independent GPU cloud providers actually is.