CoreWeave Q2 2026 earnings show $2.6B revenue and $35.6B debt
CoreWeave more than doubled quarterly revenue, but customer concentration, interest costs and a $35.6B debt load remain central risks.
By Dominic Okoye · Staff Writer
· 2 min read
CoreWeave Q2 2026 earnings showed revenue of $2.575 billion for the quarter ended June 30, up 112% from a year earlier, while total indebtedness reached $35.6 billion. The AI cloud provider also posted a $626 million net loss and expects to spend $35 billion to $39 billion on capital expenditures in 2026, keeping its expansion dependent on continued financing and demand for GPU capacity.
The results illustrate the economics facing companies building cloud infrastructure for AI workloads. Revenue growth is rapid, but the servers, networking and data-center capacity required to supply that demand require large upfront outlays. CoreWeave reported $2.624 billion in operating expenses, resulting in a $49 million operating loss. Net interest expense rose 140% year over year to $640 million, contributing to the wider net loss, according to The Register's account of the company's SEC filing.
Why did CoreWeave report a loss despite revenue growth?
Interest costs were larger than the operating loss. CoreWeave said its infrastructure buildout has left it with $35.6 billion in debt as of June 30, while it continued adding capacity. It also reported $10 billion available to borrow through revolving-credit and delayed-draw term-loan facilities. That is borrowing capacity, rather than cash on hand, and it underscores the company’s remaining access to debt financing as it executes its spending plan.
CoreWeave's reported growth was also concentrated in its installed customer base. Approximately 93% of the year-over-year revenue increase came from existing customers, with new customers supplying the balance. Three customers represented 36%, 26% and 10% of quarterly revenue, or 72% combined. Such concentration means renewal, expansion and payment decisions by a small group of buyers carry unusual weight for a company funding a large hardware buildout.
Management argues that AI use is spreading beyond model developers into enterprise software, industrial systems, financial markets and national-security work, creating recurring compute needs for training, inference, evaluation and model improvement. CoreWeave expects its managed inference service, introduced only months earlier, to reach at least a $250 million annual recurring revenue run rate by the end of 2026. That is company guidance, not reported revenue.
CoreWeave also identified AWS, Microsoft Azure and Google as larger competitors, noting that some are customers as well. The quarter therefore leaves a clear operating test: whether customer demand can keep rising fast enough to support the company’s capex program, service its interest burden and reduce reliance on a small set of large accounts. CoreWeave cautioned in its filing that it cannot predict whether its recent rate of revenue growth will continue or when it will reach positive net income.
This story draws on original reporting from The Register.