CoreWeave Adopts Customer Prepayments and Shorter Contracts to Support AI Infrastructure Growth

Image: TechCrunch · Source

CoreWeave is increasingly relying on customer prepayments, contract-backed loans, and higher-priced short-term deals to finance its rapid AI-driven expansion, while managing significant losses and debt costs.

CoreWeave, a cloud computing provider focused on powering artificial intelligence workloads, is changing its financing approach to support the surging demand for AI services. In its third-quarter earnings report, the company revealed a strategy that increasingly depends on customer prepayments, loans collateralized by specific customer contracts, and a greater emphasis on shorter-term contracts commanding higher prices.

The company reported revenue of $3.66 billion for the quarter, marking a 168% increase year-over-year, driven by strong demand for AI computing power. However, this growth was accompanied by a net loss of $709 million and hefty quarterly interest expenses of $846 million, underscoring the financial challenges of scaling AI infrastructure.

CoreWeave’s CEO Michael Intrator emphasized during the earnings call that the company will avoid unsecured debt at current high-interest yields. Instead, it plans to leverage loans secured by its substantial backlog of $132.5 billion in customer contracts, drawing funds as needed for infrastructure expansion.

The firm also anticipates receiving about $9 billion in customer prepayments and deposits over the next three months, including more than $3 billion linked to contracts for capacity that is already built, thus requiring no additional capital expenditures. This upfront cash inflow eases immediate borrowing requirements and reflects industry trends, as competitors like Nebius and IREN also rely on large prepayments from big tech companies like Microsoft to finance their capacity.

To further capitalize on the robust market, CoreWeave is increasing allocations to short-term contracts, which currently command annual prices exceeding $40 million per megawatt, compared to $20 million to $25 million per megawatt for longer four- and five-year agreements. The company raised prices approximately 25% in July and another 10% later in the quarter, with continued price growth offsetting cost increases.

In August, CoreWeave secured a $2.6 billion loan facility tailored to its strategy of funding short-term contract-backed projects. This loan, called DDTL 5.5, enables installment borrowing based on funding needs tied to these contracts. While short-term deals enhance pricing flexibility and revenue, lenders typically prefer longer-term contracts for predictable cash flow, posing a balancing act for the company's financing approach.

Despite positive reactions from Wall Street analysts to CoreWeave’s earnings, bond investors remain cautious, with some company bond yields surpassing 15%, reflecting concerns about the risks associated with heavy borrowing to fund massive data center buildouts before returns materialize.

CoreWeave's evolving strategy illustrates the high-risk, high-reward environment of the AI infrastructure boom, which hinges on ongoing, strong demand and pricing power that are yet to be fully proven over the long term. The company’s approach to financing growth through customer prepayments and contract-backed loans exemplifies adaptive measures to balance expansion with fiscal prudence in a volatile market.