Anthropic and Nscale Weigh Second $45 Billion Compute Deal

Anthropic and data center startup Nscale are in active talks for a second compute agreement valued at $45 billion, following a first partnership of the same scale announced earlier this year. If completed, the two deals would represent a combined $90 billion commitment between the AI safety company and the data center provider.

What Happened

The Tech Buzz reported on October 20, 2026, that Anthropic and Nscale are negotiating a second $45 billion compute deal, just months after the companies announced their first partnership. The terms under discussion, the timeline for signing, and the structure of the new agreement have not been disclosed publicly. Neither Anthropic nor Nscale has issued an official statement confirming or detailing the talks as of the time of publication.

Background

Anthropic is a San Francisco-based AI safety company founded in 2021 by Dario Amodei, Daniela Amodei, and other former OpenAI researchers. The company develops the Claude family of large language models and has positioned AI safety as a central pillar of its research and commercial operations.

Nscale is a data center and cloud infrastructure startup focused on providing GPU compute capacity to AI companies. The company has grown rapidly alongside broader industry demand for high-density compute resources required to train and run large AI models.

The first Anthropic-Nscale partnership, also reported at $45 billion in value, was disclosed earlier in 2026. At the time, it ranked among the largest single compute procurement agreements publicly reported in the AI industry. A second deal of identical scale would double the financial scope of the relationship between the two companies.

Scale and Industry Context

Commitments of this size reflect the capital intensity that has come to define frontier AI development. Training and operating large language models at commercial scale requires sustained access to tens of thousands of high-performance graphics processing units, along with the power infrastructure, cooling systems, and network fabric to support them.

Several major AI companies have signed long-term compute agreements with data center operators and cloud providers in 2025 and 2026, as demand for GPU capacity has consistently outpaced available supply. Contracts structured over multiple years are increasingly common as companies seek to lock in infrastructure access ahead of planned model development cycles.

Anthropic has separately disclosed revenue growth of 45 percent in recent reporting periods and has been expanding its policy and safety operations alongside its commercial business. The company has also recently announced the creation of the Anthropic Public Benefit Fund, a wholly owned entity to which it plans to donate a portion of company shares, a move that drew public comment from Elon Musk on social media.

What the Deal Would Cover

The specific workloads, data center locations, hardware configurations, and delivery schedules associated with the prospective second agreement have not been reported. It is not known whether the new deal would extend the geographic or technical scope of the first partnership or run concurrently with it under separate terms.

Nscale operates data centers in Europe and has been expanding its footprint to meet demand from AI customers. The company has not publicly disclosed its full client roster or the total contracted value of its infrastructure agreements.

What Happens Next

If the two companies reach agreement, a formal announcement detailing the terms and timeline of the second compute deal is expected to follow.