CoreWeave turns to customer prepayments and shorter contracts to fund AI growth

CoreWeave's third-quarter earnings show a company reshaping how it pays for growth: the AI cloud provider is increasingly asking customers to pay upfront, borrowing against specific contracts, and exploring shorter-term deals.

According to TechCrunch's reading of the results, those arrangements help fund expansion but could also lead to higher prices.

Upfront payments mean customers hand over cash before they consume compute, letting CoreWeave turn future revenue into present-day cash flow.

Borrowing against individual contracts shifts the lender's anchor from the company's overall credit to a specific project. That can make risk and return easier to match, and it turns the number of bankable contracts into an asset CoreWeave can manage in its own right.

The shorter contracts cut the other way: with less visibility past the next renewal, pricing and utilization become harder to predict, and the report notes that customers could end up paying more.

For a business whose core asset is GPU clusters, the pace of capital spending sets how quickly power, racks and chips become rentable capacity. Cash flow structure is therefore not an accounting detail but a ceiling on how fast the company can expand.

What to watch is whether these terms become standard across AI infrastructure, and whether customers still accept prepayment plus short commitments once compute supply loosens.