Apollo Nears $3.4 Billion Loan to Finance AI Chip Fund for xAI

Apollo Global Management is reportedly close to arranging about $3.4 billion of credit to back a vehicle that will buy Nvidia accelerators and lease that capacity to xAI, the AI company associated with Elon Musk. The structure — moving chips from outright sale to a lease-based deployment — aims to give xAI rapid access to training capacity while limiting near-term cash outlays and procurement timing risks. Sources say a long-standing investor in Musk’s ventures is helping put the package together, and that Valor Equity Partners has been reported as involved in arranging the transaction. The financing should be viewed alongside broader reported fundraising for xAI: industry accounts cite a roughly $20 billion round with an estimated $2 billion commitment from Tesla and participation from chip and infrastructure partners. That wider funding activity reflects how strategic investors are increasingly coupling capital with preferential compute or integration rights. The Apollo-led proposal sits at the intersection of private credit markets and the accelerating demand for specialized silicon, and it echoes other market moves tying capital to compute — for example, recent capital-and-commercial alignments between Nvidia and CoreWeave that channel funding to downstream GPU capacity. Beyond pure economics, the deal emerges amid parallel discussions about how xAI might pursue public markets, including preliminary talks linking xAI’s listing path to SpaceX, and against a backdrop of regulatory and legal scrutiny over xAI’s Grok products. Those developments could affect disclosure requirements, contractual covenants, and the commercial assurances lenders will seek. For Apollo and co-investors, the fund offers yield-bearing exposure to scarce accelerators but concentrates risk around a single vendor and tenant; lenders and sponsors will need clear collateral, resale provisions and asset-management plans to preserve recoverability. For xAI, leasing accelerators can accelerate model development while avoiding balance-sheet strain, but it could also create operational dependencies on third-party capacity agreements and the terms that govern them. Market dynamics — GPU supply, pricing, competing capacity from hyperscalers, and regulatory or litigation outcomes affecting xAI — will determine whether leased accelerators remain economically attractive. If completed quickly, the deal would illustrate how private credit and strategic capital are being marshaled to solve immediate compute bottlenecks for AI developers; observers should watch the final contract terms, the role of reported strategic backers like Tesla and Nvidia-linked providers, and any governance or cross-entity disclosure tied to SpaceX conversations.