Crusoe could be the next AI provider to test public markets as the cost of adding compute capacity keeps climbing.
Reports say the company has held preliminary IPO discussions with several Wall Street banks. No bank has been formally selected to lead a listing, keeping any deal at an early stage.
Access to capital is becoming a competitive factor for neoclouds trying to keep pace with enterprise AI demand.
From crypto infrastructure to AI capacity
Founded in 2018 as a cryptocurrency infrastructure company, Crusoe has since moved into AI infrastructure and cloud services. Operations now stretch from energy sourcing through data center development to Crusoe Cloud, placing the provider among specialists commonly grouped into the neocloud market.
Company figures released in June put contracted AI infrastructure capacity at 4.9 GW, with a development pipeline above 40 GW. Major projects include a 1.2 GW campus in Abilene and a separate 900 MW site supporting Microsoft as major cloud providers expand their AI infrastructure businesses.
AI infrastructure growth keeps demanding more capital
A $1.375 billion Series E round in October 2025 valued the company above $10 billion. Current pre-IPO fundraising discussions could bring in roughly $3 billion at a valuation around $35 billion.
JPMorgan is also advising on the pre-IPO financing, connecting the possible listing with an active effort to raise fresh capital. A completed round at the reported $35 billion valuation would represent a steep jump from the $10 billion-plus valuation set in October 2025.
Investor’s Business Daily compared a potential Crusoe listing with public neoclouds CoreWeave and Nebius.
CoreWeave and Nebius already give the sector public-market reference points, but Crusoe’s energy-to-cloud model differs from operators centered more heavily on GPU cloud capacity.
AI infrastructure providers are also gaining access to new forms of institutional financing, with Wall Street firms backing large data centers and compute projects as capital requirements rise.
Channel partners should expand infrastructure due diligence
MSPs, VARs, and systems integrators should separate financing strength from deliverable compute. Large funding rounds can support construction, but partners still need to verify how much capacity is live or contractually available and whether delivery dates match customer deployment plans.
Lower compute prices will not help customers much if capacity arrives late or workloads become difficult to move. Providers evaluating neocloud and service-provider infrastructure should confirm portability, support obligations, and exit options before placing large AI workloads with a newer operator.
Preliminary IPO talks do not call for procurement changes. Future capital could improve availability, geographic reach, or commercial terms across the neocloud market. Continued investment could make provider selection a larger part of AI infrastructure and hybrid cloud services.
Also read: Google is strengthening its Marvell partnership as hyperscalers spend more heavily on custom AI accelerators.