Anthropic Partners with Google and Broadcom to Deploy Several Gigawatts of Next-Generation TPUs

On April 6 local time, U.S. artificial intelligence (AI) technology giant Anthropic announced that it has signed a new agreement with Google and Broadcom to secure several gigawatts of computing capacity based on Google’s next-generation TPUs, with deployment expected to begin in 2027. This major expansion of computing infrastructure will power Anthropic’s cutting-edge Claude model and help the company meet the massive demand from its global customer base.

“This groundbreaking collaboration with Google and Broadcom is a continuation of our strategy to steadily expand our infrastructure: we are building the necessary computing power to meet the exponential growth in demand from our customer base, while also helping Claude lead the forefront of AI development,” said Krishna Rao, Chief Financial Officer of Anthropic. “To keep pace with our unprecedented growth, we are making our largest investment in computing resources to date.”

According to reports, Claude’s customer demand accelerated in 2026, with annual revenue exceeding $30 billion—far surpassing the approximately $9 billion recorded at the end of 2025. In February of this year, when Anthropic announced the completion of its Series G funding round, the company noted that over 500 enterprise customers were spending more than $1 million annually. Today, that number has surpassed 1,000, doubling in less than two months.

Anthropic stated that the vast majority of the new computing resources added through this partnership will be located in the United States, making this collaboration a significant expansion of the company’s November 2025 commitment to invest $50 billion in strengthening U.S. computing infrastructure. “This partnership also deepens Anthropic’s existing collaboration with Google Cloud—building on the increased TPU capacity we announced last October—as well as our relationship with Broadcom.”

Anthropic trains and runs Claude on a variety of AI hardware, including AWS Trainium, Google TPUs, and NVIDIA GPUs, which means we can select the most appropriate chips based on the workload. This platform diversity delivers higher performance and greater stability for customers relying on Claude for critical tasks.

Currently, Amazon remains Anthropic’s primary cloud service provider and training partner. Anthropic also stated that it will continue to work closely with AWS on Project Rainier. Claude remains the only cutting-edge AI model available to customers on all three major global cloud platforms: Amazon Web Services (Bedrock), Google Cloud (Vertex AI), and Microsoft Azure (Foundry).