Vertically integrated AI infrastructure operator Crusoe has announced the first close of its $3.9 billion Series F financing round. The investment elevates the company to a post-money valuation of $30.9 billion. The round was led by Atreides Management, Mubadala Capital, and Valor Equity Partners. Crusoe intends to deploy the fresh capital directly into expanding specialized hyperscale data centers designed to handle demanding AI training and inference workloads.
A primary catalyst for the valuation surge is an expansive commercial commitment from Wall Street. In early September 2026, details emerged regarding a five-year compute agreement between Crusoe and quantitative trading firm Jane Street valued at $13 billion. Jane Street is securing dedicated access to massive GPU clusters to deploy deep learning and reinforcement learning models across global market feeds. The transaction demonstrates that proprietary trading firms now require compute capacities rivaling traditional tech giants.
Beyond the lead sponsors, a wide consortium of institutional asset managers and financial technology funds participated in the round. Co-investors include Squarepoint Capital, Fidelity Management and Research, Baillie Gifford, and ARK Invest. The Robinhood Ventures Fund also took an equity position in the infrastructure provider. This investor roster highlights the strategic pivot of major financial players toward financing the foundation layer of machine intelligence directly.
Crusoe reported that its long-term contracted total contract value has surpassed $140 billion across its operations. This figure illustrates the massive forward planning required to guarantee access to chips, cooling, and electrical grid capacity. Both financial institutions and tech companies are locking in multi-year commitments to avoid severe supply bottlenecks. The aggregation of high-density computing capacity has consequently become a decisive factor for market leadership.
The arrangement marks a significant structural shift in quantitative finance. Earlier generations of quantitative trading relied on statistical heuristics requiring modest compute footprints. In contrast, training complex neural architectures capable of evaluating order-book dynamics across global venues requires industrial-scale data center capacity. Firms like Jane Street are securing these pipelines through long-term partnerships to preserve their quantitative edge.