In an SEC free writing prospectus tied to its IPO registration, SpaceX disclosed that on June 5, 2026 it entered into a Cloud Service Agreement with Google LLC for access to compute capacity. The filing states the capacity includes approximately 110,000 NVIDIA GPUs, plus CPUs, memory, and related components, and that the customer agreed to pay roughly 920 million dollars per month from October 2026 through June 2029, with capacity ramping up through September 2026 at a reduced fee. Over its full term the agreement is worth on the order of 30 billion dollars.
The filing sets out specific performance and exit terms. If SpaceX fails to deliver access to the committed amount of GPUs by September 30, 2026, then after a short grace period Google may terminate the agreement or accept the GPUs provided at a pro rata fee reduction. After December 31, 2026, either party may terminate on 90 days’ notice, and Google retains ownership of its content, AI models, and related data.
The deal landed just before SpaceX’s Nasdaq IPO and is one of several large cloud-compute contracts SpaceX disclosed in 2026. For a business reader, the striking point is that a hyperscaler with its own datacenters is renting bridge capacity from an outside supplier, a sign of how acute the AI compute crunch has become and how the economics of compute are reshaping deals across the industry.