- Anthropic has signed an agreement with Google and Broadcom for multiple gigawatts of next-generation TPU capacity, with deployment expected to begin in 2027.
- Anthropic’s annualized run-rate revenue has surpassed $30 billion as of April 2026, up from approximately $9 billion at the end of 2025.
- The number of business customers spending more than $1 million annually on Anthropic services doubled from 500 to over 1,000 in under two months following the company’s Series G announcement in February 2026.
- The majority of the new compute will be located in the United States, extending a November 2025 commitment to invest $50 billion in American AI infrastructure.
What Happened
Anthropic announced on April 16, 2026 that it has signed a new agreement with Google and Broadcom for multiple gigawatts of next-generation TPU capacity. The infrastructure is expected to begin coming online in 2027 and is intended to support Anthropic’s frontier Claude models as the company reports accelerating enterprise demand.
The deal deepens an existing Google Cloud relationship that had already been expanded through an increased TPU capacity agreement announced in October 2025, and extends Anthropic’s relationship with Broadcom.
Why It Matters
The agreement arrives as Anthropic’s customer base has grown sharply in 2026. The company’s annualized run-rate revenue has surpassed $30 billion — more than a threefold increase from the approximately $9 billion reported at the end of 2025. That growth trajectory is what Anthropic’s leadership describes as the basis for the infrastructure expansion.
The deal also extends Anthropic’s November 2025 commitment to invest $50 billion in American computing infrastructure. The vast majority of the new capacity will be sited in the United States, according to the company’s announcement.
Technical Details
Anthropic trains and runs Claude across a heterogeneous hardware stack — AWS Trainium, Google TPUs, and NVIDIA GPUs — which the company says allows workloads to be matched to the chips best suited for each task type. The new agreement adds next-generation TPU capacity, though the specific TPU generation and technical configuration are not disclosed in Anthropic’s announcement.
Amazon Web Services remains Anthropic’s primary cloud provider and training partner. Anthropic continues to collaborate with AWS on Project Rainier. Claude is available on all three major cloud platforms: AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry.
Who’s Affected
Enterprise customers who depend on Claude for production workloads are the primary beneficiaries of expanded capacity, though the new infrastructure does not come online until 2027. At the time of Anthropic’s Series G fundraising in February 2026, more than 500 business customers were each spending over $1 million on an annualized basis; as of this announcement, that figure exceeds 1,000.
Krishna Rao, CFO of Anthropic, attributed the move to demand-driven necessity: “We are making our most significant compute commitment to date to keep pace with our unprecedented growth,” Rao said in the company’s announcement. He characterized the agreement as part of what he described as a “disciplined approach to scaling infrastructure” designed to match compute supply to customer growth while advancing Claude’s development.
What’s Next
The new TPU capacity is expected to begin coming online in 2027. Anthropic has not disclosed the financial terms of the agreement, the specific TPU generation covered by the deal, or how the gigawatts of capacity will be distributed across training and inference workloads.
