CoreWeave announced a multiyear cloud infrastructure deal with Anthropic on Friday, adding the Claude maker to its roster of AI heavyweights. The agreement sends CoreWeave stock up 11% and comes just one day after Meta committed an additional $21 billion to the GPU-focused cloud provider, on top of its existing $14.2 billion deal. CoreWeave declined to disclose the Anthropic contract value but confirmed the infrastructure will start coming online later this year.
This Anthropic win solidifies CoreWeave's stranglehold on AI infrastructure. The company now claims nine of the ten leading AI model providers as customers, including Microsoft, OpenAI, and Meta. Only Elon Musk's xAI remains outside CoreWeave's ecosystem. That's remarkable market penetration for a company that was barely known outside crypto circles three years ago. CoreWeave's GPU-native architecture and Kubernetes-based platform clearly resonates with AI labs that need massive, flexible compute for training and inference.
The timing reveals how desperate AI companies are for reliable GPU capacity. Anthropic's deal follows their recent $4 billion investment from Amazon, which included commitments to use Amazon's Trainium chips. Now they're also betting on CoreWeave's NVIDIA-heavy infrastructure. This suggests Anthropic is hedging across multiple cloud providers rather than getting locked into any single vendor's hardware roadmap.
For developers, CoreWeave's near-monopoly on AI lab infrastructure could mean more consistent APIs and better price competition as they expand capacity. But it also creates a dangerous single point of failure for the entire AI ecosystem if CoreWeave stumbles.