Anthropic Signs $45 Billion Compute Deal With Nscale to Expand AI Capacity

Anthropic is continuing its aggressive expansion of AI infrastructure with a reported $45 billion agreement to secure computing capacity from Nscale, a British AI infrastructure company founded in 2024.

The deal will provide Anthropic with access to additional AI computing power through Nvidia’s next-generation Vera Rubin systems. The capacity is expected to begin supporting Anthropic’s services in late 2027, according to a source familiar with the agreement.

The agreement was first reported by Bloomberg and is expected to span six years. The computing resources will come from Nscale’s flagship data centre in West Virginia.

Anthropic Builds More Compute Capacity

Nscale has quickly expanded its role in AI infrastructure, including partnerships with major technology companies such as Microsoft. The company will provide Anthropic access to Nvidia’s Vera Rubin platform, which combines multiple specialised chips into a single AI computing system.

The additional capacity comes as Anthropic continues increasing its infrastructure investments to compete with other leading AI companies. The company has signed several major compute agreements over the past year to support the growing demand for its AI services.

Earlier, Anthropic signed a reported $10 billion agreement with AI cloud company Volta for additional computing capacity from a data centre in Norway. The company also reached a $5 billion compute-related agreement with AMD.

Anthropic’s Growing Network of Compute Partnerships

Anthropic has also expanded its relationship with other technology providers. In May, the company announced a major computing partnership with SpaceX, adding capacity from SpaceX data centres.

In April, Anthropic expanded its partnership with Amazon, gaining access to additional computing resources. The company has also strengthened relationships across the AI hardware and cloud ecosystem.

The push for more AI computing capacity is part of a broader industry race. Companies including OpenAI, Google and Meta are also investing heavily in data centres, chips and cloud infrastructure to support increasingly demanding AI models.

As AI models become larger and more complex, access to specialised computing resources has become a central part of competition among major AI developers.