CoreWeave Secures $3.1B Loan for AI Infrastructure Expansion

CoreWeave's recent closure of a $3.1 billion delayed draw term loan facility, known as DDTL 5.0, highlights the evolving nature of AI infrastructure financing. This capital raise aims to support the rapid expansion of the company's AI cloud infrastructure, particularly to fulfill contracts with two major non-investment-grade customers. The loan, maturing in approximately 5.5 years, carries an interest rate of SOFR plus 4.50%, adjusted down from the initial pricing due to strong investor demand.

The DDTL 5.0 facility is particularly significant as it is the first publicly syndicated financing vehicle based on high-performance computing infrastructure. CoreWeave noted that the transaction was “meaningfully oversubscribed,” indicating rising institutional interest in this segment of the debt market. This structure facilitates easier secondary market trading, potentially broadening access to GPU-backed financing beyond traditional private credit options.

Despite receiving a Ba2 rating from Moody's and a BB+ from Fitch—placing it below investment grade—the loan marks a growing acknowledgment of AI infrastructure as a viable asset class. Morgan Stanley and Mitsubishi UFJ Financial Group served as joint lead arrangers and bookrunners for the transaction.

This financing follows CoreWeave's earlier achievement of an $8.5 billion loan facility, DDTL 4.0, which was supported by more stable, investment-grade customer contracts. That facility received stronger ratings due to its solid collateral and contracted cash flows, allowing borrowers to access a larger portion of the total loan amount initially. The contrast between the two facilities highlights the split in AI infrastructure credit: stronger contracts can secure lower-cost debt, while weaker-rated contracts require higher yields and come with increased risk.

CoreWeave's aggressive financing strategy reflects a broader trend in the AI infrastructure sector, where companies are increasingly using customer contracts and costly compute equipment as collateral for substantial debt packages. This new loan brings CoreWeave's total debt and equity financing for the year to over $20 billion, underscoring the intense capital demands within the AI market.

As the company works to meet the rising demand for AI compute from both enterprise and hyperscale customers, the delayed draw structure of the DDTL 5.0 aligns financing with deployment schedules and the operational lifespan of GPU infrastructure. Such strategies are becoming essential for companies aiming to manage the heavy upfront costs associated with GPUs, power, and data center capabilities before realizing revenue from long-term contracts.

As institutional investors become more familiar with the unique characteristics of AI infrastructure, the demand for diverse financing solutions is likely to persist. The differing nature of CoreWeave's loan facilities illustrates the increased scrutiny and varying cost of capital that companies face based on the credit quality of their customer contracts. The emerging landscape points to a growing sophistication in how AI infrastructure is financed, with implications for future market dynamics.