Take-or-Pay Long-Term ContractsNarrow moat

CoreWeave (CRWV) — moat facet

Customers pay for the capacity whether they use it or not — as binding as cloud contracts get.

The structure that makes CoreWeave's backlog valuable is the take-or-pay long-term contract: customers commit to pay for reserved GPU capacity over multi-year terms whether or not they fully use it. This is a genuinely favorable arrangement that converts uncertain, spot-market demand into contracted, predictable revenue, shifts utilization risk onto the customer, and — crucially — provides the reliable cash flows that let CoreWeave finance its GPU fleet with debt. Lenders will advance money against contracted revenue that a spot business could never support, so the take-or-pay contracts are the linchpin that makes the entire debt-financed build-out possible. They are a real source of durability and the backbone of the business model.

Deferred revenue, customer prepayments ($ billion)$4.1BDec 2024$8.2BDec 2025$9.7BJun 2026CoreWeave 10-K FY2025 and 10-Q Q2 2026 (current plus non-current)
Customers have paid $9.7 billion ahead of service: the take-or-pay contracts come with cash, not only promises.

But a contract is only as good as the counterparty behind it, and this is where the take-or-pay structure's strength meets CoreWeave's concentration risk. The commitments are worth their headline value only if the customers remain willing and able to honor them across the full term, and CoreWeave's contracts are dominated by a few enormous counterparties — including OpenAI, a heavily loss-making company whose multi-billion-dollar commitments depend on its own uncertain, ongoing funding. A take-or-pay contract with a financially fragile counterparty carries credit risk that a headline backlog number obscures; and even strong counterparties may seek to renegotiate terms if the market shifts, technology changes the economics, or their own strategies evolve toward building rather than renting. The take-or-pay model is a real and valuable structure that underpins CoreWeave's financing and revenue visibility, genuinely stronger than spot demand. But an investor should weigh it as commitments that are only as reliable as the concentrated, and in places fragile, counterparties behind them — durable on paper, and exposed to the credit and renegotiation risk that concentration in a few large, entangled customers — ten clients now above $1B each1 — inevitably brings.

Moat trajectory: Widening

Widening. Take-or-pay commitments convert spot demand into contracted revenue and underpin the debt financing — genuinely valuable. But a contract is only as good as its counterparty, and CoreWeave's are concentrated and in places fragile.

The number that tests this moat
Reported
Clients committed above $1B each
10

Customers pay for the capacity whether they use it or not — as binding as cloud contracts get, with ten billion-dollar clients compounding quarterly. Take-or-pay is only as strong as the counterparty's solvency and patience; a single renegotiation would set the precedent for all ten.

Source: CoreWeave Q2 2026 release ↗
⚠ Threats to the moat
References
  1. ReportedTen clients now above $1B each.
    CoreWeave Q1/Q2 2026 earnings releases — Q2 revenue ~$2.5B (+111%); revenue backlog $99.4B (Mar 2026, from $66.8B at end-2025); 2026 capex guided $31–35B; ~$28B of financing raised in 12 months; quarterly interest expense >$500M; ten clients >$1B each — Q1-Q2 2026 · publ. 2026 · source ↗
Sources
Generated September 23, 2026