Anthropic Secures Multi-Gigawatt Compute Expansion via New Google and Broadcom Agreement

Anthropic has announced a significant expansion of its computing power through a new strategic agreement with Google and Broadcom. The deal secures multiple gigawatts of next-generation Tensor Processing Unit (TPU) capacity expected to come online in 2027. This massive infrastructure boost is designed to support the development of future frontier Claude models and meet surging global demand.

The expansion comes as Anthropic experiences historic financial growth in the enterprise sector. The company's run-rate revenue has now surpassed $30 billion, a staggering increase from the $9 billion reported at the end of 2025. This growth is mirrored by their customer base, with over 1,000 business clients now spending more than $1 million annually on Claude services.

Strategic partnership with Broadcom is central to this scaling effort. Broadcom will serve as a key intermediary for custom silicon manufacturing, helping Anthropic access the hardware necessary for unprecedented AI workloads. Most of the new compute resources will be located within the United States to fulfill a prior $50 billion commitment to domestic infrastructure.

Anthropic continues to maintain a diversified hardware and cloud strategy to ensure high performance and resilience. While Amazon Web Services remains the primary cloud provider and training partner, the company runs Claude across Google TPUs, AWS Trainium, and NVIDIA GPUs. This approach allows the firm to match specific AI tasks to the most efficient chips available.

The move to secure hardware years in advance highlights the increasing difficulty of obtaining specialized compute resources. By establishing this long-term pipeline, Anthropic aims to maintain its position at the frontier of AI development. The agreement ensures that Claude remains accessible across all major cloud platforms, including AWS, Google Cloud, and Microsoft Azure.

Alongside this infrastructure news, the company is launching the Anthropic Institute to study the societal impacts of powerful AI. They also recently signed an MOU with the Australian government to focus on safety research. These initiatives complement the technical scaling by addressing the governance and safety challenges of next-generation systems.