Anthropic Signs $40 Billion Compute Deal With SpaceX for Claude AI
Anthropic has agreed to pay Elon Musk’s SpaceX nearly $45 billion over the next three years for computing resources to support its Claude artificial intelligence software, according to a securities filing.
The financial terms of the Anthropic SpaceX compute deal publicly not known until SpaceX filed its S-1 registration statement with the U.S. Securities and Exchange Commission this week ahead of a planned Nasdaq listing under the ticker SPCX.
A Deal Built on Mismatched Demand
Anthropic expected to pay $1.25 billion per month through May 2029, with capacity ramping in May and June 2026 at a reduced fee. The agreement gives Anthropic access to 300 megawatts of compute capacity from the Colossus 1 data center near Memphis, Tennessee, housing roughly 220,000 GPUs.
The logic behind the Anthropic SpaceX compute deal is straightforward. Colossus 1 was built to train Grok, Musk’s AI assistant, but Grok hasn’t filled it. Anthropic, on the other hand, is on track for more than $40 billion in revenue and simply needed more compute.
Anthropic is also expanding beyond Colossus 1 to Colossus 2, scaling up on Nvidia GB200 capacity throughout June. Either side can terminate the contract with 90 days’ notice, and SpaceX has indicated it expects to enter additional similar services contracts going forward.
Anthropic is also reportedly preparing for an IPO targeted for October 2026, with market discussions placing a possible valuation near $1 trillion. The compute agreement expected to feature prominently in investor analysis as public markets weigh the true cost of running frontier AI at scale.
