Anthropic and Nscale in talks for $45 billion compute deal

By Splitfeed EditorialLast updated Oct 20, 2026, 01:50 UTC

On October 19, 2026, Bloomberg News reported that Anthropic and Nscale, a cloud computing company backed by Nvidia, are in negotiations over a compute agreement valued at approximately $45 billion. An SEC filing from Nscale cited by at least one outlet pins the figure more precisely at $44.6 billion. The talks are described as a second major deal between the two companies, suggesting an existing and deepening commercial relationship. The deal's scale is reported to be a factor in Nscale's own IPO valuation negotiations, meaning the contract serves both companies' financial positioning simultaneously. On the same day, Reuters reported that OpenAI is seeking to accelerate a $30 billion fundraising round, a move described as a response to Anthropic's approaching IPO.

The size of the reported figure places this among the largest single compute commitments disclosed in the AI industry to date. It reflects how dependent frontier AI labs have become on securing guaranteed infrastructure at enormous scale, and how that infrastructure spending is now directly entangled with capital markets. The fact that both Anthropic and OpenAI are making major financial moves on the same day, with the two events framed as competitive, points to a race dynamic in which access to compute and investor capital are inseparable. What remains genuinely in dispute is whether these deals represent sustainable business models or inflated pre-IPO positioning.

The Pro-AI camp has produced no published reactions to this story yet. Based on their typical positions, they would likely frame the deal as validation of Anthropic's commercial trajectory and a sign that AI infrastructure investment is maturing into a serious industry. The Anti-AI camp has responded with sharp skepticism about the companies involved rather than the specific deal. Hart Cunningham on Bluesky raised concerns about Claude eventually acting against human interests, while clnzjay described AI companies broadly as parasites profiting from journalists' work without compensation, and NonsenseBot called for the elimination of Anthropic and its peers entirely. The Middle Ground camp, as represented by CHICAGO NEWS BENCH, has focused on financial risk, flagging that if the IPOs from OpenAI and Anthropic fail to perform, related stocks could face collateral damage.

The clearest near-term test of the arguments on all sides will be the Anthropic IPO itself, whenever it is filed and priced. Whether the company's valuation holds up in public markets, and whether the Nscale deal is treated by investors as an asset or a liability, will go a long way toward settling the question of whether these compute commitments reflect genuine demand or pre-IPO leverage.