AI labs and infrastructure teams that need large scale GPU clusters for training and inference.

Crusoe is an energy first cloud provider offering on demand GPU clusters, including NVIDIA GB200, B200, H200, and AMD MI300 series hardware, plus managed fine tuning through Crusoe Intelligence Foundry.

Fully hosted across Crusoe owned data centers including Abilene Texas, Iceland, and Wyoming, accessed via cloud console or API.

Founded in 2018 and pivoted heavily into AI GPU cloud from around 2022 onward, raising a large Series E in late 2025.

Pairs stranded or renewable energy sources with GPU data centers to offer competitively priced large scale AI compute.

Offers standard cloud primitives plus Kubernetes and Slurm support for distributed training and inference jobs.

Commercial SaaS requiring an account, usage billed per GPU hour, workloads run on Crusoe owned physical infrastructure.