Anthropic has agreed to use all compute capacity at SpaceX’s Colossus 1 data center, adding more than 300 megawatts of new capacity and over 220,000 Nvidia GPUs within the month. The partnership gives Claude a major infrastructure boost at a time when demand for AI coding, enterprise agents, and high-end model access has strained usage limits across frontier AI platforms.
The Colossus 1 cluster includes Nvidia H100, H200 and GB200 accelerators, giving Anthropic access to a GPU mix built for large language models, multimodal systems and high-throughput inference workloads. For Claude users, the impact is immediate. Claude Code’s five-hour rate limits are doubling across Pro, Max, Team and seat-based Enterprise plans. Peak-hours limit reductions are also being removed for Claude Code on Pro and Max accounts, while API rate limits for Claude Opus models are rising substantially.
The deal turns compute scarcity into a user-facing product change. Claude Code has become one of Anthropic’s most important growth products, especially among developers who use it for software engineering, code review, debugging and agentic workflows. Higher rate limits give paid users more room to run longer sessions without hitting throttles during heavy demand periods.
The SpaceX agreement adds to Anthropic’s already aggressive compute expansion. The company also has an up to 5 GW agreement with Amazon, a 5 GW Google and Broadcom capacity deal scheduled to begin coming online in 2027, a Microsoft and Nvidia partnership that includes $30 billion of Azure capacity, and a $50 billion American AI infrastructure investment with Fluidstack.
The pattern reflects a wider market reality: frontier AI companies are now competing on data centers, power access, chips and inference capacity as much as model quality. The same pressure has been visible across the broader AI capex boom in chips, where demand for semiconductors, networking gear and power infrastructure has become one of the strongest equity-market themes.
Anthropic’s SpaceX deal also arrives as semiconductor stocks sit inside a broader AI chip supercycle. More Claude capacity means more demand for GPUs, data-center power and operating efficiency, reinforcing the link between AI software growth and the hardware supply chain behind it.
Anthropic has also expressed interest in working with SpaceX on multiple gigawatts of orbital AI compute capacity. That idea remains exploratory, but it shows how far the infrastructure race has moved from ordinary cloud procurement. Frontier model providers are now looking at land, grid access, cooling, power availability, chip supply and even space-based data centers as strategic constraints.
Claude users on paid coding and API products are getting higher limits now, and Anthropic is adding one of the largest available GPU clusters to its compute stack. Colossus 1 gives the company more room to serve developers and enterprise customers while AI demand keeps pushing every major lab toward bigger power deals, larger GPU deployments and tighter control over the infrastructure behind each model response.
The post Anthropic Taps SpaceX’s Colossus 1 As Claude Usage Limits Jump appeared first on Crypto Adventure.
