CoreWeave hits $100B backlog as enterprise AI adoption accelerates

Enterprise AI workloads are moving beyond training into production, and CoreWeave is capturing the infrastructure spend.

CoreWeave has secured nearly $100 billion in contracted revenue backlog as enterprises shift AI workloads from training into inference and production, with financial services emerging as a major growth vertical.

"Enterprise adoption is accelerating, customer commitments continue to increase, and the expanding platform positions CoreWeave to support a broader range of AI workloads," management said in the company's latest business update.

The company recorded more than $40 billion in new commitments during the first quarter, bringing total contracted backlog to nearly $100 billion. Ten customers have each committed at least $1 billion. Financial services alone is approaching $10 billion in revenue backlog, supported by expanded commitments from existing clients like Jane Street and new additions such as Hudson River Trading. Additional demand is emerging from physical AI and spatial computing, with robotics and autonomous driving firms including World Labs, PhysicsX and Sunday Robotics selecting CoreWeave's platform.

The backlog provides multi-year revenue visibility for CoreWeave as it competes with hyperscalers and rivals like Nebius for enterprise AI workloads. The company is expanding beyond GPU rental into a full-stack cloud platform with CPU, storage, networking and developer tools, introducing products such as CoreWeave Omni that let customers deploy its cloud stack within their own data centers. Nebius, a competitor also benefiting from enterprise AI adoption, has reported growing demand from manufacturing, energy and pharmaceutical companies for its full-stack AI offerings.

CoreWeave's platform expansion targets a broader set of enterprise requirements. The company launched a Trust Center to help organizations meet security and compliance standards for AI deployments. It also introduced Flex Reservation and Spot pricing models, along with cross-cloud connectivity solutions, to give customers more flexibility in how they consume AI infrastructure. These additions signal a shift from a pure GPU provider to a more comprehensive cloud platform capable of supporting the full lifecycle of AI workloads.

The enterprise AI infrastructure market is expanding as organizations move beyond experimental deployments into production environments. CoreWeave's addressable market now includes inference, agentic workloads and enterprise-scale production — all of which require significant computing power. The company's customer concentration, however, remains a risk: with 10 clients each committing at least $1 billion, the loss of any single large customer could meaningfully affect revenue. Capacity details, including total megawatt buildout and data center expansion plans, have not yet been disclosed.

CoreWeave shares, trading on the Nasdaq under CRWV, have benefited from the AI infrastructure buildout as enterprises accelerate production deployments. The nearly $100 billion backlog suggests the market may not have fully priced in the duration and scale of committed revenue. The company's ability to deliver on its contracted commitments while competing against hyperscalers with larger balance sheets will determine whether this backlog translates into sustained margin expansion and shareholder returns.

This article is for informational purposes only and does not constitute investment advice.