Crusoe in talks to raise $3B at reported $30B valuation

The neocloud arms race just got a fresh valuation marker. Bloomberg reports that Crusoe is in active talks to raise about $3 billion, with investors expecting the round to value the company in the $30 billion range. That would roughly triple the $10 billion-plus mark set by its Series E last year, which was co-led by Valor Equity Partners and Mubadala Capital and drew in NVIDIA, Founders Fund and Fidelity, among others.

The specifics worth internalizing are less about the headline number and more about what backs it. Crusoe said in June that it holds contracts for 4.9 gigawatts of computing power and has more than 40 gigawatts in its total project pipeline. Meta is already under contract to buy roughly 1.6 gigawatts of capacity across two Crusoe sites in Childress, Texas, and Warrenton, Missouri, per Bloomberg's earlier reporting. Oracle is on the customer list too.

For anyone tracking where the AI infrastructure money is actually landing, this is the shape of it: private companies stringing together power, land and GPU supply into multi-gigawatt facilities and selling that capacity forward to a small handful of hyperscalers and model labs. Crusoe launched in 2018 as a cryptocurrency business and has re-tooled itself into one of the more visible names in the neocloud category. A tripling in under a year, if the round closes at the reported mark, is a pretty stark read on how investors are pricing that pivot.

The honest caveat is that the reporting is single-sourced and explicitly tentative. Bloomberg notes a final valuation has not been set. Reasonable people can argue whether $30 billion for a company whose revenue is anchored to a handful of anchor tenants is a fair mark or a peak-of-cycle print. What the reporting doesn't give you is round composition, expected close timing, or how much of that 40-plus gigawatt pipeline is genuinely site-controlled versus early-stage.

If the number holds, the interesting downstream effects are on the other neoclouds still raising against Crusoe as the comp, on Crusoe's ability to lock in more Meta- or Oracle-scale offtake before the next hyperscaler capex cycle turns, and on when a public-market exit starts looking like the obvious next step.