On October 23, 2025, Anthropic announced plans to expand its use of Google Cloud technologies to as many as one million Tensor Processing Units (TPUs). The company said the expansion is worth tens of billions of dollars and is expected to bring well over a gigawatt of compute capacity online in 2026.
Why this matters
Anthropic framed the increase in compute as a response to rapidly growing enterprise demand and as necessary infrastructure for more extensive testing, alignment research and responsible large‑scale deployment. The company currently serves more than 300,000 business customers, and the number of large accounts—defined as customers representing more than $100,000 in run‑rate revenue each—has grown nearly sevenfold in the past year.
Statements from Google Cloud and Anthropic
Thomas Kurian, CEO of Google Cloud, said Anthropic’s decision to significantly expand TPU usage reflects the strong price‑performance and efficiency Google’s TPU platform has demonstrated. Google highlighted continued innovation and capacity improvements across its TPU family, including its seventh‑generation TPU, Ironwood.
Krishna Rao, Chief Financial Officer of Anthropic, said the expansion will help the company scale the compute it needs to advance the Claude model and meet exponentially growing demand from customers ranging from Fortune 500 firms to AI‑native startups.
Multi‑platform compute strategy
Anthropic emphasized a diversified compute approach that leverages three chip platforms: Google’s TPUs, Amazon’s Trainium, and NVIDIA’s GPUs. According to the announcement, this multi‑platform strategy enables continued progress on Claude while preserving industry partnerships.
The company also reiterated its ongoing partnership with Amazon, which remains Anthropic’s primary training partner and cloud provider. Anthropic and Amazon are collaborating on Project Rainier, described as a massive compute cluster with hundreds of thousands of AI chips distributed across multiple U.S. data centers.
Continued investment
Anthropic stated it will continue to invest in additional compute capacity to keep its models and capabilities at the forefront of AI research and deployment.
