- CoreWeave posts record Q2 revenue of $1.21 billion and a backlog of $30.1 billion, driven by surging AI cloud demand.
- Net loss widens to $290.5 million amid soaring operating expenses; CEO flags infrastructure and power constraints.
What happened: Q2 revenue beats expectations as backlog hits record
CoreWeave posted US$1.21bn in revenue for Q2 2025, surpassing market forecasts and reporting a record US$30.1bn backlog of contracted business. According to Capacity Media, this reflects strong AI workload demand from customers including OpenAI. For independent confirmation, see CoreWeave’s official investor release and Reuters’ earnings coverage.
The company operates dozens of GPU-focused data centres across the US and Europe, offering vertical-optimised compute, storage, and managed services. However, net loss widened to US$290.5m, while operating expenses surged to US$1.19bn, underlining the heavy capital and operational costs of rapid expansion.
Also read: Nokia secures deal with Nscale for AI cloud data centre networking
Also read: Kao Data partners with CBRE to expand its AI data centre footprint
Why it’s important
CoreWeave’s performance underscores its pivotal role in meeting escalating AI compute needs—but also raises concerns about its financial model. Can the company sustain rapid expansion as infrastructure costs balloon and power limitations emerge? Amid AI fervour, the widening gap between top-line gains and net losses risks investor confidence if growth fails to convert to profitability.
Furthermore, reliance on a narrow customer base and looming debt obligations challenge CoreWeave’s resilience. As power constraints tighten, even robust demand may falter—prompting scrutiny on whether AI-driven growth can be both scalable and economically sound.
