Key Takeaways
- • Anthropic has committed to a multi-gigawatt TPU capacity deal with Google and Broadcom, worth tens of billions of dollars, with infrastructure expected to come online starting in 2027.
- • Anthropic will deploy up to 1 million TPU chips as part of a broader $200 billion spending commitment to Google Cloud over five years, representing the company's largest compute infrastructure investment to date.
- • Anthropic's annualized run-rate revenue has surpassed $30 billion as of May 2026, more than triple the $9 billion reported at the end of 2025, driven by accelerating customer demand for Claude models.
- • The partnership with Google and Broadcom expands Anthropic's multi-chip strategy, which also includes AWS Trainium and NVIDIA GPUs, enabling workload optimization and operational resilience across heterogeneous hardware platforms.
- • The majority of new compute capacity will be sited in the United States, aligning with Anthropic's November 2025 commitment to invest $50 billion in American computing infrastructure.
Summary
Anthropic has secured a landmark infrastructure partnership with Google and Broadcom for multiple gigawatts of next-generation TPU (Tensor Processing Unit) capacity in a deal valued at tens of billions of dollars. According to Anthropic's official announcement, the company will deploy up to 1 million TPU chips as part of the agreement, with infrastructure expected to become operational beginning in 2027. This commitment represents Anthropic's most significant compute infrastructure investment to date and underscores the intensity of resource competition in frontier AI development.
The partnership builds on Anthropic's broader $200 billion five-year spending commitment to Google Cloud announced earlier, establishing Google as the primary cloud infrastructure provider for the company's Claude model training and deployment. Krishna Rao, Anthropic's Chief Financial Officer, characterized the deal as essential to meeting exponential demand growth: "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."
The timing of this deal reflects extraordinary business momentum at Anthropic. The company's annualized run-rate revenue has surpassed $30 billion as of May 2026, representing a more than threefold increase from approximately $9 billion at year-end 2025. Enterprise adoption has accelerated sharply: the number of business customers spending over $1 million annually on Anthropic services has doubled to over 1,000 in less than two months since the company's February Series G fundraising announcement. This rapid scaling of customer spending justifies the massive compute commitments now being locked in.
Anthropic maintains a diversified hardware strategy, training and running Claude across AWS Trainium chips, Google TPUs, and NVIDIA GPUs. This multi-platform approach allows the company to optimize workloads for specific hardware while mitigating supply chain and operational risk. The partnership reflects deliberate infrastructure planning rather than dependency on a single vendor. Notably, the vast majority of new compute capacity will be located in the United States, continuing Anthropic's stated commitment from November 2025 to invest $50 billion in domestic computing infrastructure, with potential implications for U.S. technology sovereignty and competitiveness in AI infrastructure development.
The partnership builds on Anthropic's broader $200 billion five-year spending commitment to Google Cloud announced earlier, establishing Google as the primary cloud infrastructure provider for the company's Claude model training and deployment. Krishna Rao, Anthropic's Chief Financial Officer, characterized the deal as essential to meeting exponential demand growth: "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."
The timing of this deal reflects extraordinary business momentum at Anthropic. The company's annualized run-rate revenue has surpassed $30 billion as of May 2026, representing a more than threefold increase from approximately $9 billion at year-end 2025. Enterprise adoption has accelerated sharply: the number of business customers spending over $1 million annually on Anthropic services has doubled to over 1,000 in less than two months since the company's February Series G fundraising announcement. This rapid scaling of customer spending justifies the massive compute commitments now being locked in.
Anthropic maintains a diversified hardware strategy, training and running Claude across AWS Trainium chips, Google TPUs, and NVIDIA GPUs. This multi-platform approach allows the company to optimize workloads for specific hardware while mitigating supply chain and operational risk. The partnership reflects deliberate infrastructure planning rather than dependency on a single vendor. Notably, the vast majority of new compute capacity will be located in the United States, continuing Anthropic's stated commitment from November 2025 to invest $50 billion in domestic computing infrastructure, with potential implications for U.S. technology sovereignty and competitiveness in AI infrastructure development.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.