CoreWeave’s interest expense reaches $640M, up 2.4 times year-over-year

CoreWeave has a revenue story that most startups would kill for. It also has an interest bill that would make a leveraged buyout specialist wince.

The AI cloud infrastructure company reported Q2 2026 interest expense of $640 million, up 140% from $267 million in the same quarter a year ago. Revenue more than doubled to $2.575 billion, up from $1.212 billion year-over-year. The GAAP net loss, meanwhile, widened to $626 million compared to $290 million in Q2 2025.

The debt pile keeping things interesting

CoreWeave ended Q2 2026 with roughly $30 billion in long-term debt on its balance sheet. During the quarter alone, it raised over $10 billion in new debt and capital, including a $3.1 billion term loan and a $1 billion investment from Jane Street.

The company’s full-year 2026 capital expenditure guidance now sits at $35 to $39 billion, most of it going toward NVIDIA GPU-heavy data centers.

On the revenue side, the case for optimism is real. The company’s revenue backlog reached $104 billion, up 246% year-over-year, driven by demand from hyperscalers, AI labs, and enterprise customers. The adjusted operating income of $128 million shows the underlying compute business can generate positive cash flow from operations before accounting for the debt servicing.

What comes next, and why the numbers get bigger before they get smaller

CoreWeave guided for Q3 2026 interest expense of $860 to $940 million. The company is projecting its interest bill will grow by roughly 40% in a single quarter.

Q3 revenue guidance of $3.45 to $3.6 billion suggests the top line continues to scale rapidly. But with interest expense projected to approach or exceed $900 million in Q3, the gap between revenue growth and debt servicing costs remains a defining tension in CoreWeave’s financial story.

CoreWeave’s origins add a layer of context worth noting. The company started as a crypto mining operation before pivoting aggressively into AI cloud infrastructure as GPU demand from the machine learning world overtook crypto mining economics. The company went from mining Ethereum to becoming one of the primary GPU rental platforms for frontier AI labs, a transformation that took just a few years.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.