- Lead. CoreWeave reported Q2 revenue of $2.58 billion, up 112% year over year and ahead of the $2.56 billion consensus estimate, as surging demand for GPU compute pushed the company’s revenue backlog to $104 billion — up 246% from a year earlier.
- Fact. The AI cloud provider lifted its full-year 2026 revenue guidance to $12.4–$13.2 billion and raised planned capital expenditure to $35–$39 billion, a figure that dwarfs the company’s entire public valuation at the time of its March 2026 IPO.
- Stake. CoreWeave shares surged roughly 13.5% in after-hours trading on August 11, adding to gains that have made the company one of the best-performing new public listings of the year and a bellwether for institutional confidence in dedicated AI infrastructure spending.
CoreWeave reported second-quarter 2026 revenue of $2.58 billion on August 11, beating analyst estimates of $2.56 billion and posting 112% year-over-year growth as hyperscalers and enterprise customers continued to expand their GPU compute contracts. The result drove shares up approximately 13.5% in after-hours trading, extending a rally that has made the Nvidia-GPU neocloud one of the most closely watched names in AI infrastructure.
Adjusted EBITDA came in at $1.5 billion, representing a 59% margin and offering the first clearest evidence of operating leverage in the company’s model since it went public. The net loss narrowed relative to forecasts, with adjusted earnings per share of -$1.14 compared with the -$1.41 loss analysts had expected. The company held $6.9 billion in cash at quarter end after spending $9.4 billion on capital expenditure during the quarter alone.
Backlog and Demand Signal
The figure that drew the most market attention was the revenue backlog: $104 billion as of quarter end, up 246% from a year earlier and calculated without the more than $25 billion in new customer commitments signed in the third quarter that are not yet reflected in the total. CEO Mike Intrator told analysts that Q2 “demonstrated exceptional execution with operating leverage materializing in results as the company scales infrastructure and customer demand across sectors.”
CFO Nitin Agrawal highlighted that margin-accretive ancillary businesses — storage, CPU, networking and software — now exceed $400 million in annualised revenue. CoreWeave’s managed inference revenue, meanwhile, grew from $1 million to $100 million in annualised rate within months and is expected to reach $250 million by year-end, reflecting customers shifting from reserved GPU capacity to consumption-based inference.
CapEx Escalation and Infrastructure Context
CoreWeave raised its full-year capital expenditure range to $35–$39 billion — an extraordinary commitment for a company that went public less than six months ago. The company now controls 1.5 gigawatts of active power capacity after adding 500 megawatts in the quarter alone, with 4.2 gigawatts under contracted supply. It is targeting 8 or more gigawatts by 2030. That trajectory puts it in direct competition — and partnership — with the hyperscalers whose AI infrastructure financing is reshaping data centre construction globally.
For the third quarter CoreWeave guided revenue of $3.45–$3.60 billion, implying continued sequential acceleration. Full-year adjusted operating income is expected to reach $960 million to $1.15 billion on revenues of $12.4–$13.2 billion, with an exit annualised revenue rate of $18.5–$19.5 billion by December. The results sent semiconductor-adjacent stocks broadly higher in after-hours trading, reinforcing the market’s view that demand for AI compute infrastructure remains well ahead of supply.