SpaceXAI has agreed to give Anthropic access to Colossus 1, one of the world’s largest AI supercomputers, to relieve growing demand on Claude’s infrastructure. Anthropic will use the cluster to increase capacity for Claude Pro and Claude Max subscribers as user demand strains existing resources.
The arrangement provides Anthropic access to a cluster with more than 220,000 NVIDIA GPUs, including H100, H200, and next‑generation GB200 accelerators. Colossus 1 is built to handle AI training, fine‑tuning, inference, high‑performance computing, large language models, multimodal systems, scientific simulations, and generative AI workloads.
The deal highlights escalating competition for AI compute as labs compete for chips, power, and data‑center infrastructure needed to train and run larger models. Providers are racing to secure the hardware and energy capacity required to scale next‑generation systems on the timelines customers expect.
Beyond immediate capacity needs, Anthropic has signaled interest in a longer‑term collaboration with SpaceXAI to develop orbital AI compute — multiple gigawatts of capacity in space — arguing that terrestrial power, land, and cooling may be insufficient to meet future demand on required schedules. The partnership pairs SpaceXAI’s substantial compute resources with Anthropic’s operational use to support Claude’s paying customers.
Anthropic is also expanding Claude’s scope beyond chat into coding, enterprise workflows, and agentic applications. On the same day, the company announced new Claude Managed Agents features — dreaming, outcomes, and multiagent orchestration — intended to help agents retain memory of prior work, verify outputs, and coordinate more complex tasks with reduced human intervention.
Edited by Estefano Gomez.