Anthropic’s new compute agreement with SpaceX gives the AI company access to all compute capacity at SpaceX’s Colossus 1 data center, adding more than 300 megawatts of capacity and more than 220,000 NVIDIA GPUs within the month. While the immediate impact is higher capacity for Claude users, the deal also stands out as a major utilization win for SpaceX’s AI infrastructure ambitions.
The agreement comes as AI companies continue to lock down large-scale compute wherever they can find it. For Anthropic, the SpaceX capacity will directly improve availability for Claude Pro and Claude Max subscribers and support broader growth across Claude Code and the Claude API. For SpaceX, the partnership brings a major external AI workload into Colossus 1, giving the facility an immediate customer for a large block of capacity.
SpaceX Deal Adds More Than 300 Megawatts of Capacity
Under the agreement, Anthropic will use all compute capacity at SpaceX’s Colossus 1 data center. The company says this gives it access to more than 300 megawatts of new capacity within the month, including more than 220,000 NVIDIA GPUs.
The added capacity is expected to directly improve service for Claude Pro and Claude Max subscribers, and it also gives Anthropic another major compute resource at a time when AI model providers are competing aggressively for power, GPUs, and data center access.
There is also a broader strategic angle for SpaceX, as Industry chatter has suggested that AI data center capacity tied to Elon Musk’s companies, including xAI, may not have been fully utilized at certain points. Anthropic’s agreement gives SpaceX a major customer for Colossus 1 capacity, which could help demonstrate demand for large-scale AI infrastructure connected to Musk’s broader compute ecosystem.
SpaceX has also been linked to future space-based data center ambitions, and Anthropic says it has expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute capacity. No timeline or technical details were provided, but the agreement gives both companies a practical starting point for a much larger compute relationship.
Compute Demand Moves Beyond Traditional Cloud Deals
The SpaceX agreement joins several other compute arrangements already underway at Anthropic. The company has an agreement with Amazon for up to 5 gigawatts of capacity, including nearly 1 gigawatt of new capacity by the end of 2026. It also has a 5-gigawatt agreement with Google and Broadcom that will begin coming online in 2027, a strategic partnership with Microsoft and NVIDIA that includes $30 billion in Azure capacity, and a $50 billion investment in American AI infrastructure with Fluidstack.
Claude runs on a range of AI hardware, including AWS Trainium, Google TPUs, and NVIDIA GPUs. The SpaceX agreement adds another large NVIDIA GPU deployment to that mix, expanding Anthropic’s access to AI infrastructure outside the usual hyperscaler channels.
The deal could also play well for SpaceX as it looks to show that its infrastructure can support more than launch, satellite, and communications businesses. If SpaceX moves toward an IPO, a large AI compute customer using hundreds of megawatts of capacity would be a useful proof point for investor discussions around future revenue lines and infrastructure demand. That remains speculative, but the Anthropic agreement provides SpaceX with a concrete example of external demand for its compute capacity.
International Expansion Remains Part of the Plan
Some of Anthropic’s capacity expansion will also take place outside the United States. Enterprise customers, especially those in regulated sectors such as financial services, healthcare, and government, increasingly need in-region infrastructure to meet compliance and data residency requirements. Anthropic’s collaboration with Amazon includes additional inference capacity in Asia and Europe.
Anthropic says it will be selective about where it adds international capacity, focusing on democratic countries with legal and regulatory frameworks that can support infrastructure investment at this scale. The company is also weighing supply chain security, including the hardware, networking, and facilities needed to support AI compute.
Anthropic recently committed to covering any increases in consumer electricity prices caused by its data centers in the United States. As it expands internationally, the company is exploring ways to extend that commitment to new jurisdictions and work with local leaders on investments in communities that host its facilities.
For Anthropic, the SpaceX agreement is a near-term capacity boost for Claude. For SpaceX, it is a large-scale AI infrastructure customer at a time when compute demand, data center utilization, and future orbital compute plans are becoming more important to the company’s broader story.
Amazon