April 8, 2026 Anthropic has signed a new agreement with Google and Broadcom to significantly expand the compute capacity powering its Claude AI models. The deal includes access to roughly 3.5 gigawatts of compute and will come online starting in 2027.
The expansion builds on an earlier agreement with Google Cloud and reflects a sharp rise in demand for Anthropic’s models, particularly from enterprise customers. The company said the majority of the new infrastructure will be based in the United States as part of a broader $50 billion commitment to domestic compute investment.
The partnership centres on increased use of Google’s tensor processing units (TPUs), specialised AI chips designed for training and running large models. The scale of the agreement, measured in gigawatts rather than data centre count, underscores how quickly compute has become a primary constraint in AI development.
Anthropic’s growth has accelerated alongside this infrastructure push. The company reported a $30 billion revenue run rate, up from $9 billion at the end of 2025. It also said more than 1,000 enterprise customers are now spending over $1 million annually on its services.
To support that expansion, Anthropic recently raised $30 billion in a Series G funding round, valuing the company at $380 billion. The funding and infrastructure commitments together signal a focus on scaling capacity fast enough to meet demand from large customers.
The timing is notable given broader scrutiny. The U.S. Department of Defense has flagged Anthropic as a potential supply-chain risk, even as the company continues to deepen its partnerships with major U.S. technology providers.
For Anthropic, the strategy is straightforward: secure enough compute to keep pace with usage growth and maintain competitiveness in a market where access to infrastructure increasingly determines product capability.
