Anthropic has reportedly agreed to spend about $45 billion on artificial intelligence computing capacity from British infrastructure provider Nscale, in one of the largest AI infrastructure commitments disclosed to date. According to reporting cited by Data Center Dynamics and Bloomberg, the six-year arrangement would give Anthropic access to roughly 460 megawatts of capacity at Nscale’s Monarch Compute Campus in West Virginia. The reported agreement highlights the scale of infrastructure required by leading AI developers as they expand increasingly compute-intensive models and services.
A Major Capacity Commitment
The contract is expected to cover systems built around Nvidia’s next-generation Vera Rubin platform, with capacity scheduled to begin coming online in late 2027. Nscale has not publicly confirmed the reported Anthropic agreement, and Data Center Dynamics said it had contacted the company for comment, making the transaction a reported deal rather than a formally announced partnership. If completed on the terms reported, the commitment would strengthen Anthropic’s access to large-scale computing resources over several years.
Nscale Expands in West Virginia
The capacity would be supplied from Nscale’s Monarch Compute Campus, a large data center development in West Virginia that the company acquired in March 2026. The site spans about 2,250 acres and has been designed for potential capacity of up to 8 gigawatts, with an initial 2-gigawatt phase expected to come online during the first half of 2028. The development is intended to support large AI workloads and is becoming a key part of Nscale’s effort to establish itself as a major provider of high-density computing infrastructure.
Microsoft Also Backs the Campus
Nscale has already secured another major customer for the West Virginia project, with Microsoft committing to 1.35 gigawatts of capacity at the campus. That deployment is expected to use Nvidia Vera Rubin NVL72 systems, reinforcing the site’s role as a large-scale hub for advanced AI hardware. The presence of multiple large customers could help Nscale spread development costs while strengthening the commercial case for expanding the campus toward its planned capacity.
Anthropic’s Broader Compute Expansion
The reported Nscale agreement comes as Anthropic continues to increase its access to cloud and data center infrastructure across several providers. Data Center Dynamics has reported that the AI company has committed to more than 10 gigawatts of servers through cloud providers and has also pursued additional data center leases with developers in the United States. Its previously reported infrastructure relationships include Google, Amazon Web Services, Akamai, CoreWeave, and Fluidstack, reflecting a strategy built around securing diversified sources of computing power.
Nscale Prepares for the Public Markets
The timing of the reported Anthropic contract is also significant for Nscale as the company prepares for a potential initial public offering in the United States. Recent reports have indicated that the British AI infrastructure company could seek to raise around $3 billion through the offering, while earlier reporting said it had accumulated tens of billions of dollars in contracted business. A long-term agreement with a customer of Anthropic’s scale could strengthen Nscale’s growth narrative as investors assess the company’s revenue visibility, capital requirements, and ability to deliver major infrastructure projects.
The reported $45 billion agreement between Anthropic and Nscale illustrates how access to power, data centers, and advanced chips has become a central competitive factor in the artificial intelligence industry. For Anthropic, the proposed six-year commitment would add another substantial source of computing capacity as it scales its models and services, while Nscale would gain one of the sector’s most prominent AI developers as a major customer. The deal remains subject to formal confirmation, but its reported size underscores the accelerating financial and infrastructure demands behind the global race to build and operate frontier AI systems.
Source: Bloomberg