Anthropic Accelerates AI Infrastructure Investments with $35 Billion Lambda Deal

Anthropic is rapidly scaling up its investments in the infrastructure required to develop its artificial intelligence models. According to a source familiar with the matter, cited by Reuters, the company behind Claude has reportedly signed a $35 billion cloud computing agreement with Lambda, an infrastructure provider backed in part by Nvidia.

Texas Data Center to Power Claude’s Growth

The agreement centers on a data center currently under construction in Nueces County, Texas. Developed by Hut 8, the facility is designed to deliver 350 MW of power capacity and will house Nvidia equipment, providing Anthropic with the computational resources necessary for its operations.

Lambda is set to deploy the American manufacturer’s processors, enabling Anthropic to train and operate its models, including Claude, as well as its various applications.

Nvidia at the Heart of the Compute Race

This deal underscores Nvidia’s growing role in shaping the AI ecosystem. According to the Wall Street Journal, the American tech giant may itself hold the lease for the Texas data center. Lambda, in which Nvidia has invested, would then install the chips procured from the manufacturer.

Anthropic is not alone in securing massive contracts to meet surging computing demands. The company recently signed a six-year agreement with Nscale, valued at $45 billion, to access computing capacity in West Virginia. This infrastructure is expected to provide up to 460 MW of power and will rely heavily on Nvidia’s Vera Rubin chips.

The New Frontier: Infrastructure Over Models

For Anthropic, the immediate priority is securing sufficient computing power to support the rapid growth of its services. Demand is particularly rising around Claude Code, its tool designed for developers.

These multi-billion-dollar contracts signal a shift in the AI landscape: competition among major players is no longer defined solely by model performance, but increasingly by access to data centers, advanced chips, and the energy required to power them.