Anthropic Signs Historic Compute Deal to Fuel Frontier AI Growth
Anthropic partners with Google and Broadcom for 3.5GW of compute capacity to power Claude AI, amid a surge to $30B in run rate revenue.
A Massive Infrastructure Bet on AI’s Future
Anthropic, the high-profile artificial intelligence research lab, has solidified a major partnership with Google and Broadcom to drastically scale its computing power. Announced on Monday, the agreement is designed to provide the necessary hardware to keep pace with the exploding demand for its Claude AI models.
The deal centers on expanding Anthropic's access to Google Cloud’s Tensor Processing Units (TPUs), custom-built, high-performance chips designed specifically for machine learning. This move builds upon a previous agreement signed in October 2025 and represents the most ambitious infrastructure commitment in the company’s history.
Scaling Up to 3.5 Gigawatts
While Anthropic’s official blog post was light on specific technical metrics, a recent Broadcom filing with the Securities and Exchange Commission (SEC) revealed the scale of the commitment: 3.5 gigawatts of compute capacity.
This massive infrastructure project is expected to:
Announced on Monday, the agreement is designed to provide the necessary hardware to keep pace with the exploding demand for its Claude AI models.
- Go online by 2027, providing a long-term roadmap for model development.
- Support a $50 billion investment in U.S.-based computing infrastructure.
- Maintain U.S. leadership in frontier AI development, as the majority of the hardware will be housed domestically.
“This groundbreaking partnership... is a continuation of our disciplined approach to scaling infrastructure,” stated Krishna Rao, CFO of Anthropic. “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.”
Explosive Economic Growth
The infrastructure deal comes at a time of "unprecedented growth" for the San Francisco-based firm. Despite recently being labeled a supply-chain risk by the U.S. Department of Defense, Anthropic’s financial trajectory remains steep.
The company recently reported a staggering run rate revenue of $30 billion, a massive leap from the $9 billion recorded at the end of 2025. This growth is largely driven by enterprise adoption; Anthropic now boasts more than 1,000 business customers who spend at least $1 million annually.
Anthropic also recently finalized a $30 billion Series G funding round, which skyrocketed the company’s valuation to $380 billion, cementing its status as one of the most valuable private technology companies in the world.
The Hardware Arms Race
By deepening ties with Google and Broadcom, Anthropic is securing its place in the global "compute race." As AI models become more complex, the ability to access high-end chips and the electricity required to run them has become the primary bottleneck for the industry. This 3.5-gigawatt deal ensures that Anthropic will not be left behind as it competes with rivals like OpenAI and Meta for AI supremacy.
