Google has signed one of the largest cloud infrastructure agreements in recent memory: a contract to pay SpaceX $920 million per month for AI compute capacity, running from October 2026 through June 2029. The arrangement gives Google access to approximately 110,000 Nvidia GPUs, along with CPUs, memory, and supporting infrastructure housed in SpaceX data centers.
The deal was publicly disclosed in the week before SpaceX’s anticipated IPO, which aims to raise approximately $75 billion at a $1.75 trillion valuation. Google is a longtime investor in SpaceX, and its holdings are expected to exceed $100 billion after the public offering.
Why Google needed outside capacity
According to a company statement, the arrangement serves as bridge capacity to handle unexpectedly high demand for Gemini Enterprise, Google’s agent platform for business customers. A Google representative stated that this is a short-term, timely agreement to ensure bridge capacity to meet surging customer demand for the Gemini Enterprise agent platform.
The deal is structured to scale gradually: pricing ramps up through September 2026 at reduced rates before reaching the full $920 million monthly figure in October. Both parties retain the right to terminate the agreement with 90 days notice after December 31, 2026.
A pattern forming in AI infrastructure
This deal is not isolated. Anthropic signed a similar compute agreement with SpaceX in May 2026, roughly half the size of Google’s arrangement. The pattern suggests that major AI companies are running out of capacity at existing cloud providers and are turning to unconventional suppliers, including one previously known primarily for rockets and satellites.
SpaceX’s entry into AI infrastructure hosting is a significant strategic pivot. Its track record in large-scale hardware deployment and Starlink’s global network give it a credible claim to become a serious compute provider. Google and SpaceX are also reportedly exploring orbital data centers together.
What it means for AI in Latin America and globally
For businesses in Latin America evaluating enterprise AI platforms, Gemini Enterprise’s capacity expansion means more reliable availability and potentially faster regional deployment. At $920 million per month, Google will spend over $10 billion annually on this single infrastructure agreement through 2029.
The broader message is that AI at scale requires physical infrastructure investment of a magnitude that few companies can sustain. The leaders that can secure compute at scale today are building structural advantages that will be difficult for later entrants to overcome.
For sales teams tired of cold leads, slow customer responses, and manual processes, Dapta is the ultimate tool.
Dapta is the leading platform for creating AI sales agents specifically designed to increase inbound lead conversion. Respond to your leads in less than a minute with voice AI and WhatsApp that converts.
If you want your team to sell more while AI handles the complex stuff, you have to try it.
The Google-SpaceX compute deal is a signal of where AI infrastructure investment is flowing: not just into model development, but into the raw physical capacity needed to run AI products at global scale. The companies that lock in this capacity now are positioning themselves for the AI-native enterprise market of the next decade.
