top of page

Anthropic and Nscale Near Second $45 Billion Cloud Capacity Agreement, Sources Say

Writer: Andrej Botka
Andrej Botka
7 hours ago
1 min read

Both companies are reportedly close to finalizing a second $45 billion contract to secure cloud processing capacity, according to people familiar with the discussions who cautioned terms could still shift. The talks, first reported by Bloomberg, remain fluid and no binding paperwork has been disclosed.


The potential arrangement would follow a similar announcement between the two firms roughly five weeks earlier, underlining a rapid cadence of large-scale capacity commitments by the AI start-up. Company spokespeople did not immediately offer comment when contacted for this report.


Investors have reacted unevenly to the flurry of high-dollar infrastructure commitments. Some shareholders worry that locking up vast sums for compute resources could compress returns, particularly as dominant hardware suppliers post large profits that concentrate power in the sector. A market analyst who tracks AI infrastructure noted that such repeat agreements can stoke concern among retail and institutional investors alike, even as builders argue the spending is necessary to scale advanced models.


Both Anthropic and Nscale are pursuing initial public offerings. Anthropic is still widely expected to pursue a listing before year’s end, while Nscale faces investor resistance over a proposed $30 billion valuation, according to recent reporting. An IPO adviser working with cloud firms said pushback on price tags is common in hot markets and can force companies to adjust timelines or target numbers.


If completed, another blockbuster contract would reinforce the pattern of big-ticket capacity deals shaping the AI market and could intensify competition for data-center resources. Observers say the outcome will hinge on final terms, and sources emphasize that negotiations can move in either direction until contracts are signed.

 
 
 

Recent Posts

See All
Fix Decision Processes Before You Expect AI To Do It

A lot of organizations blame artificial intelligence when projects don't change how work actually gets done. But the real obstacle is how decisions are made — and who’s responsible for them. When the

 
 
 

Comments


Subscribe here to get our latest posts

© 2026 by The StartupsCentral. 

  • X
bottom of page