According to documents filed by SpaceX with the U.S. Securities and Exchange Commission (SEC), Google has signed a long-term computing power procurement agreement with SpaceX. Under the terms of the deal, Google will pay $920 million monthly to SpaceX from October 2026 through June 2029 in exchange for large-scale artificial intelligence compute resources. The filings indicate that this partnership involves approximately 110,000 NVIDIA GPUs, along with supporting CPUs, memory, and other related computing infrastructure. As stipulated in the contract, either party may terminate the collaboration starting in 2027 upon providing 90 days’ notice to the other. Regarding the rationale behind signing the agreement, Google stated it was primarily driven by surging customer demand for its generative AI platform, Gemini Enterprise. A spokesperson for Google Cloud said in a statement that while Google Cloud has maintained a longstanding cooperative relationship with SpaceX, this arrangement represents a phased capacity expansion intended to provide transitional compute support for Gemini Enterprise. The current market growth rate for the platform’s services exceeds company expectations. Notably, this marks neither the first time nor an isolated instance of SpaceX generating revenue by leasing out computing assets. According to previously disclosed documents, AI firm Anthropic also entered into a similar cooperation with SpaceX. Under that agreement, Anthropic agreed to continue paying $1.25 billion per month to utilize compute resources provided by SpaceX’s Colossus data center, with the term extending until May 2029. As SpaceX advances its highly anticipated initial public offering (IPO) plans, such compute-leasing businesses have gradually become one of its significant income sources. In relevant disclosures, SpaceX noted that opening temporarily idle computational resources to external customers allows the company to improve infrastructure utilization rates and generate additional profits. Meanwhile, this collaborative approach preserves flexibility for future reallocation of these resources according to internal business needs. Against the backdrop of continuously rising demand for generative AI and tight global supply of high-end GPU resources, large-scale purchasing and leasing of computing power is emerging as a critical competitive tool among tech giants and AI enterprises. SpaceX is leveraging its massive data-center infrastructure to carve out new commercial growth opportunities under these conditions.