Anthropic strikes multibillion-dollar Google chip deal to expand Claude capacity

Anthropic has signed a multibillion-dollar deal with Google to secure more computing power for its Claude chatbot, the company said on April 3, 2026, in a move that highlights the escalating race among artificial intelligence developers to lock in the infrastructure needed to train and run advanced models.

According to the company, the agreement will give Anthropic access to up to 1 million of Google’s AI chips and is expected to bring well over a gigawatt of capacity online in 2026. The deal adds another large-scale commitment to a market in which access to chips and data-center power has become a central competitive advantage.

Capacity becomes a strategic priority

Anthropic said the chips will support Claude, its AI assistant that competes with OpenAI’s ChatGPT and other business-focused systems. The company has been expanding its infrastructure footprint as demand for coding, enterprise and other high-volume AI applications continues to grow.

Google’s specialized AI chips are known as Tensor Processing Units, or TPUs. Anthropic said its systems also run on Nvidia chips and on Amazon’s cloud computing division, which remains its primary cloud provider.

A larger bet on enterprise AI

The agreement comes as major AI companies continue to spend heavily on the hardware and power required to support model development and deployment. Anthropic, founded in 2021 by former OpenAI leaders, said last month that its valuation had reached $183 billion after it raised another $13 billion in investments.

The company has positioned Claude as a tool for business customers, particularly in coding and productivity workflows. The new chip deal suggests Anthropic is preparing for further growth in those markets, where compute availability can shape product performance and rollout speed.

Google deepens its role in AI infrastructure

For Google, the deal extends its role not only as a direct competitor in AI products but also as a supplier of the underlying infrastructure that powers rival systems. Google has been promoting its TPUs as a core part of its AI stack, and the Anthropic agreement gives that strategy a prominent external customer.

The arrangement also reflects the broader shift in the AI sector toward long-term supply commitments. Rather than relying on spot purchases of computing capacity, leading model developers are increasingly signing large agreements to secure access to chips, cloud services and energy-intensive data-center capacity.

What to Watch

The key question is how quickly Anthropic can bring the new capacity online and whether the added compute translates into faster product development or broader availability for Claude users. Investors and competitors will also be watching whether the deal encourages more large-scale infrastructure agreements across the AI industry in the months ahead.