Anthropic said on Monday it has signed a new agreement with Google and Broadcom to secure additional processing power to support its Claude AI models, as demand for its systems continues to surge.

The revised arrangement will expand Anthropic’s use of Google Cloud’s tensor processing units, the company’s specialised AI chips, building on an October 2025 deal that provided more than a gigawatt of compute capacity.

Anthropic said in a blog post that the new compute capacity is expected to come online in 2027.

While the company did not disclose detailed figures about the expansion, a recent Broadcom filing with the United States Securities and Exchange Commission indicates the deal covers 3.5 gigawatts of compute power.

Most of the infrastructure will be located in the U.S. and forms part of Anthropic’s broader $50 billion commitment to invest in U.S. compute capacity, the company added.

“This groundbreaking partnership with Google and Broadcom is a continuation of our disciplined approach to scaling infrastructure: we are building the capacity necessary to serve the exponential growth we have seen in our customer base while also enabling Claude to define the frontier of AI development,” Krishna Rao, CFO of Anthropic, said in the press release. “We are making our most significant compute commitment to date to keep pace with our unprecedented growth.”

Anthropic has experienced a sharp surge in demand for its Claude models in recent months, driven largely by enterprise adoption, even after the United States Department of Defense classified the company as a supply-chain risk.

The company also recently secured $30 billion in a Series G funding round, bringing its valuation to $380 billion.