CoreWeave just locked in a multi-year, $6.8 billion agreement with Anthropic to power the Claude family of AI models. The Information first reported on the deal, which was officially announced on April 10 and sent CoreWeave’s stock surging 11% in a single session.
The timing is striking. This announcement landed just one day after CoreWeave revealed a $21 billion commitment from Meta. Two landmark infrastructure deals in 48 hours.
What the Deal Covers
Under the agreement, CoreWeave will provide dedicated cloud infrastructure to support both the development and deployment of Anthropic’s Claude models. Key details:
- GPU supply: The deal secures access to NVIDIA Vera Rubin GPUs, the latest generation of AI training silicon
- Phased rollout: Compute capacity comes online later this year and scales over time
- US-based: All data center capacity will be located in the United States
- Expansion potential: The initial deal is structured with room to grow
CoreWeave declined to disclose the exact financial terms, but multiple outlets have pegged the deal at approximately $6.8 billion, making it one of the largest single commitments for dedicated AI infrastructure ever signed.
Why This Matters
With the addition of Anthropic, nine of the ten leading AI model providers now run workloads on CoreWeave’s platform. That’s a remarkable concentration of the AI industry’s most compute-hungry companies on a single infrastructure provider.
For Anthropic specifically, this deal signals serious scaling ambitions. Training and serving frontier AI models requires enormous GPU clusters, and locking in multi-year capacity with CoreWeave gives Anthropic supply certainty at a time when GPU access remains one of the biggest bottlenecks in AI development.
CoreWeave’s total contract backlog now exceeds $66.8 billion, supporting the company’s 2026 revenue guidance of $12-13 billion. For a company that went public just recently, that kind of demand pipeline is extraordinary.
The Bigger Picture
The AI infrastructure arms race is accelerating. The hyperscalers (AWS, Azure, GCP) still dominate cloud computing overall, but CoreWeave has carved out a niche as the go-to GPU cloud provider for AI-native companies that need dedicated, high-performance compute without the overhead of building their own data centers.
Anthropic previously relied heavily on partnerships with Amazon Web Services and Google Cloud. Adding CoreWeave to the mix diversifies Anthropic’s infrastructure supply chain, which is a smart move when demand for AI compute continues to outpace supply across the industry.
What stands out here is the speed at which CoreWeave is stacking these deals. Meta’s $21 billion commitment on Tuesday, Anthropic’s $6.8 billion on Thursday. The company is rapidly becoming the infrastructure backbone for frontier AI development.
For AI practitioners and companies building on Claude, this deal means Anthropic is investing heavily in the compute needed to scale its models and services. More infrastructure capacity translates to better availability, lower latency, and the ability to train increasingly capable models.
More details are available in the original report from The Information.