Crusoe

Crusoe (Crusoe Energy Systems) is a vertically integrated "AI factory company" that develops and operates gigawatt-scale, energy-first AI data centers. Founded in 2018 by Chase Lochmiller and Cully Cavness and headquartered in San Francisco with major operations in Denver, Crusoe began by capturing stranded and flared natural gas to power compute and has since become one of the fastest-growing AI infrastructure developers in the United States.

In October 2025 the company closed an oversubscribed $1.375 billion Series E at a $10 billion valuation, co-led by Valor Equity Partners and Mubadala Capital with participation from NVIDIA, Founders Fund, and Fidelity; by July 2026 it was reported to be in talks to raise around $3 billion in a pre-IPO round at a valuation near $30 billion, roughly triple the Series E mark. Crusoe's flagship project is the 1.2-gigawatt campus in Abilene, Texas — the first site of the Stargate program for OpenAI and Oracle, engineered to host up to roughly 400,000 NVIDIA GB200 GPUs — which secured about $11.6 billion in financing and brought its first phase online in September 2025.

The company's power pipeline has grown more than fourfold to over 45 gigawatts, including a newly announced 1.8-gigawatt campus in Wyoming. Crusoe serves AI labs, hyperscalers, and enterprises that need rapidly deployed, power-secured, NVIDIA-based training and inference capacity, positioning it as a leading independent builder of the physical layer of AI.

Crusoe is the one company in this group that most buyers cannot actually buy from, and that distinction is the point. Where CoreWeave, Nebius, Lambda and RunPod rent GPUs to end customers, Crusoe develops and operates the campuses — land, power, cooling, shell — largely for hyperscalers and frontier labs at gigawatt scale, which puts it upstream of its apparent competitors and in some cases supplying them.

Its March 2026 announcement of a 900MW Abilene campus for Microsoft, taking the wider Abilene footprint to roughly 2.1GW alongside the Stargate site, is the clearest illustration: the counterparties are the largest technology companies in the world, not engineering teams. The differentiator is energy strategy rather than software — the flare-gas origins matured into a general thesis that power availability, not chip supply, is the binding constraint on AI capacity, which is why that development pipeline is measured in gigawatts rather than GPUs and is the number to watch.

Buyer caveat: there is no self-serve tier and no published price list, engagements are project-scale and multi-year, and the model concentrates exposure in construction financing and long-dated power agreements. Best fit: organisations procuring dedicated capacity by the megawatt.

Everyone else is a customer of Crusoe's customers.