Anthropic will gain access to over 300 megawatts of compute capacity, equivalent to more than 220,000 NVIDIA GPUs, from SpaceX’s Colossus 1 data center within the month, a rapid scaling achieved through a partnership with the space exploration company. This increase in resources is immediately improving access to Anthropic’s AI models, as the company doubles Claude Code’s five-hour rate limits for Pro, Max, Team, and Enterprise plans and removes peak hour restrictions for Pro and Max accounts. These changes, effective immediately, aim to improve the experience for dedicated customers. Beyond leveraging hyperscalers like Amazon, Google, and Microsoft, Anthropic is also making direct capital investments, including a $50 billion commitment to American AI infrastructure with Fluidstack.
SpaceX Partnership Adds 300 Megawatt Compute Capacity
This substantial increase, equivalent to over 220,000 NVIDIA GPUs, represents a rapid scaling of resources and a departure from reliance solely on traditional hyperscalers for AI infrastructure. The addition of SpaceX’s capacity will directly benefit Claude Pro and Claude Max subscribers, addressing growing demand for the AI models. The partnership with SpaceX is not Anthropic’s only effort to improve user experience; the company simultaneously announced increased usage limits for its Claude Code service and the Claude API. Anthropic trains and operates Claude on a diverse range of hardware, including AWS Trainium, Google TPUs, and NVIDIA GPUs, and is actively seeking further capacity expansion. Anthropic has also expressed interest in collaborating with SpaceX to develop multiple gigawatts of orbital AI compute capacity, signaling a long-term vision for distributed and potentially space-based AI processing. The company is also prioritizing international expansion, with plans to provide in-region infrastructure for enterprise customers facing data residency requirements, and is committed to responsible infrastructure development, including covering potential electricity price increases caused by its data centers in the US.
Increased Claude Code & Opus API Rate Limits
Anthropic is responding to growing demand for its Claude models with a strategy focused on increasing compute capacity and improving API access for developers and dedicated users. This influx of resources, equivalent to over 220,000 NVIDIA GPUs, is already translating into tangible improvements for users of Claude Code and the Claude API. The company has eliminated peak hour restrictions on Claude Code for Pro and Max accounts, providing consistent access regardless of demand. These changes, effective immediately, are aimed at improving the experience of using Claude for its most dedicated customers. The most significant API adjustments involve Claude Opus models, with substantially raised rate limits detailed in a recent announcement. This commitment to expanding compute infrastructure extends beyond immediate capacity gains; Anthropic has also invested $50 billion in American AI infrastructure with Fluidstack, signaling a long-term strategy of direct capital investment rather than sole reliance on third-party providers. Anthropic is also prioritizing international expansion, focusing on partnerships with democratic countries to ensure secure and compliant data residency for enterprise customers.
We’ve signed an agreement with SpaceX to use all of the compute capacity at their Colossus 1 data center.
