SpaceXAI has leased its Colossus 1 supercomputer to the AI startup Anthropic to boost the compute capacity of its Claude models [1, 2].
The agreement marks a significant shift in the AI infrastructure landscape, as Elon Musk now provides the essential hardware needed for a direct competitor to scale its operations. The deal positions SpaceXAI as a major hyperscale AI infrastructure provider [3, 5].
The multibillion-dollar deal [4] involves the use of the Colossus 1 system, which is located at the SpaceX campus in Texas [3]. This partnership provides Anthropic with the vast compute power required to train and run its large language models [3].
This collaboration comes despite a history of public friction between Musk and the startup. In March 2024, Musk said, "Anthropic is evil" [6]. Musk has previously criticized the company's approach to AI, often labeling competitors as "woke" [1].
Anthropic said the deal will allow the company to double Claude Code's five-hour rate limits for Pro users [7]. By securing this capacity, the startup can reduce bottlenecks in its processing power and accelerate the deployment of new features for its user base.
SpaceXAI is leveraging its existing data-center facilities to create a new revenue stream from AI infrastructure [4]. The move allows the company to monetize its hardware investments while providing the raw processing power that remains the most constrained resource in the AI race [5].
“"Anthropic is evil"”
This partnership illustrates the growing divide between the AI software layer and the infrastructure layer. While Musk maintains a public ideological conflict with Anthropic, the commercial necessity of compute power has created a pragmatic interdependence. SpaceXAI is transitioning from a company that simply builds AI to one that controls the physical means of production for the entire industry.
