Crusoe raises $3.9bn Series F at $30.9bn valuation for AI infrastructure

Crusoe has closed a $3.9bn Series F co-led by Atreides, Mubadala Capital and Valor Equity, valuing the AI infrastructure company at $30.9bn.

A server rack with glowing blue data cables and green indicator lights stands in the foreground, with an aisle of blurred server racks extending into the distance under bright overhead lights in a data center.

Crusoe has announced the initial closing of a $3.9 billion Series F funding round at a post-money valuation of $30.9 billion, in what ranks among the largest single raises in AI infrastructure to date. The oversubscribed round was co-led by Atreides Management, Mubadala Capital, and Valor Equity Partners, with participation from Founders Fund, GIC, NVIDIA, Qatar Investment Authority, Radical Ventures, TPG, and a lengthy roster of additional institutional and strategic backers.

The Denver-based company says it has accumulated more than $140 billion in total contracted value across its platform, serving a customer base it describes as spanning AI-native companies, hyperscalers, frontier model developers and enterprises. Named cloud customers include Cognition, Figure, and Perplexity. The release also notes that OpenAI trained what the company refers to as its first artificial general intelligence model at a Crusoe-built campus in Abilene, Texas.

The platform and what the capital funds

Crusoe positions itself as vertically integrated across the full AI infrastructure stack: originating and managing power directly, building and operating data centres, manufacturing modular "Crusoe Spark" units in the US, and running a cloud platform that includes infrastructure-as-a-service, serverless fine-tuning, and a managed inference product. The company reports more than 6 GW of gross contracted capacity, with 1 GW described as delivered and operational today.

The managed inference product, launched late last year, has reportedly crossed $100 million in contracted annual recurring revenue. Crusoe claims its inference engine, powered by proprietary "MemoryAlloy" technology, delivers up to 9.9 times faster time-to-first-token and five times higher throughput than vLLM. Those benchmark figures are vendor-reported and have not been independently verified by this publication, though the company says it holds the top-ranked inference speed position on the Artificial Analysis benchmarking service and has been designated an NVIDIA Exemplar Cloud. Gartner placed Crusoe in the Visionary quadrant of its Magic Quadrant for the relevant category, according to the release.

The new capital will be deployed into scaling existing programmes and constructing what Crusoe calls "AI factories": large-scale vertically integrated campuses alongside modular Spark units. The company employs more than 1,800 people across five countries and says it is actively hiring.

Market context and competitive positioning

The AI infrastructure build-out has become one of the defining capital allocation stories of this investment cycle. Hyperscalers continue to commit record capex to GPU clusters and associated power and cooling infrastructure, while a cohort of specialist operators, including CoreWeave, Lambda Labs, and Crusoe itself, have attracted sovereign wealth and strategic investment as an alternative to on-hyperscaler compute.

Crusoe's energy-first differentiation sets it apart from competitors that treat power as a site-selection afterthought. The company has historically sourced stranded or otherwise underutilised energy, and its current strategy spans grid, battery, nuclear, thermal, and renewable partnerships. That approach carries genuine operational complexity but, if executed at scale, could offer a structural cost advantage as power constraints tighten across conventional data-centre markets in the US and Europe.

The valuation of $30.9 billion places Crusoe in a small cohort of private infrastructure companies that have achieved hyperscaler-tier capital commitments without a public listing. NVIDIA's direct participation as an investor is notable: it signals both conviction in the platform and the strategic importance to NVIDIA of cultivating well-capitalised, GPU-intensive customers. Regulatory scrutiny of large AI infrastructure deals has so far been limited, but the involvement of sovereign wealth funds from the Gulf and Singapore may draw review under US national-security frameworks as the company scales internationally.

The Series F follows a period of rapid commercial expansion. Cloud bookings are reported to have grown more than 20 times year-on-year in 2026 to date, though absolute revenue figures were not disclosed. Investors and customers will be watching for continued delivery against the 6 GW contracted pipeline, named enterprise wins beyond the current disclosed set, and a prospective public markets timeline.