Crusoe in Talks to Raise $3B at a $30B Valuation

Crusoe, the data-center builder that supplies computing power to some of the biggest names in artificial intelligence, is in talks to raise about $3 billion in a funding round that could value the company at roughly $30 billion, Bloomberg reported on July 2. That would nearly triple the roughly $10 billion valuation the startup carried as recently as October — a jump achieved in under nine months, and a fresh marker of how much capital is chasing the physical backbone of the AI boom.

Crusoe has quietly become one of the most important names in AI infrastructure. It builds and operates the large data centers that train and run modern AI systems, and its customer list reads like a who's who of the industry: it has supply arrangements touching Microsoft, Oracle, OpenAI and Meta Platforms. Crusoe is a lead builder on OpenAI's Stargate project, including a 1.2-gigawatt cluster in Abilene, Texas, and has data centers under construction for Meta in Texas and Missouri. The company has raised roughly $2.77 billion to date, most recently a $1.375 billion Series E, according to its own disclosures.

What sets Crusoe apart is how it builds. Rather than pouring concrete site by site, the company assembles data centers from prefabricated modules that ship pre-loaded with server racks, power management and cooling, orchestrated by its Command Center software. It also runs its own AI-optimized public cloud offering Nvidia and AMD accelerators. In March 2026 it opened a manufacturing facility to produce those modular components at scale, with first shipments expected this quarter — and a $3 billion war chest would let it pour that money straight into faster build-out.

The round is still being negotiated, and a final valuation has not been set; the reports did not name the new investors, though late-stage rounds like this typically draw existing backers such as Nvidia and Salesforce Ventures. Even so, the sheer velocity of the re-rating — from $10 billion to a reported $30 billion in months — underscores how investors are now treating AI data-center capacity as one of the scarcest and most valuable assets in technology, on par with the models it runs.