SpaceX Secures Google AI Compute Deal Worth US$920m A Month Ahead Of IPO

KUALA LUMPUR,June, 2026 – SpaceX has signed a major multi-year cloud services agreement with Google, marking another major step in the rocket company’s expansion beyond space launches and satellite communications into artificial intelligence infrastructure.

Under the agreement, Google will pay SpaceX US$920 million per month from October 2026 to June 2029 to secure computing capacity, according to a regulatory filing cited by Reuters. Capacity will be ramped up through September at a reduced fee before the full monthly payment begins.

The computing capacity covered by the deal includes around 110,000 Nvidia GPUs, as well as CPUs, memory and other related components. These resources are expected to support Google’s growing AI workload as demand for enterprise AI services continues to accelerate.

The agreement comes as Google works to meet rising demand for Gemini Enterprise, its agentic AI platform. A Google Cloud spokesperson told Business Insider that the arrangement is a short-term bridge to meet stronger-than-expected customer demand for the platform.

The deal also strengthens SpaceX’s AI infrastructure narrative ahead of its highly anticipated US stock market debut. Reuters reported that SpaceX is targeting a US$75 billion raise at an expected valuation of around US$1.75 trillion, potentially making it one of the largest initial public offerings in history.

The Google agreement follows another major compute deal involving Anthropic. SpaceX previously reached an agreement to provide Anthropic with access to computing power from its Colossus 1 facility in Memphis, Tennessee, which houses more than 220,000 Nvidia processors.

Together, SpaceX’s disclosed compute access deals with Google and Anthropic are worth roughly US$26 billion annually and more than US$70 billion in aggregate, assuming the contracts run until their scheduled end dates.

The size of the Google deal shows how intense the race for AI computing power has become. As artificial intelligence models become more advanced, technology companies need access to enormous amounts of GPU capacity to train, deploy and run AI products for businesses and consumers.

For Google, the agreement provides additional computing capacity at a time when demand for AI services is growing quickly. Enterprise customers are increasingly using AI agents, cloud-based tools and model-powered applications, creating pressure on cloud providers to expand infrastructure faster.

For SpaceX, the deal creates a new and significant revenue stream outside its traditional businesses. The company is already known for rocket launches, Starlink satellite internet and space infrastructure, but the latest agreement shows that it is increasingly positioning itself as a major player in AI computing.

This shift is important because SpaceX is preparing for a public listing at a time when investors are closely watching how companies can monetise artificial intelligence. AI infrastructure has become one of the most valuable areas in global technology, with demand for chips, data centres and energy capacity continuing to rise.

SpaceX’s move into cloud computing also reflects a broader strategy to turn its technical infrastructure into a business platform. The company said in its filing that such arrangements allow it to monetise unused compute capacity while retaining flexibility to reallocate the capacity for internal initiatives if needed.

That flexibility could be valuable for SpaceX as it develops future AI-related systems for aerospace, robotics, satellite networks and internal operations. By selling capacity to external customers, the company can generate revenue while still maintaining access to infrastructure for its own long-term ambitions.

The agreement also includes termination provisions. Reuters reported that if SpaceX does not provide access to the agreed number of GPUs by September 30, Google may terminate the deal after a one-month grace period or accept a smaller number of GPUs with a proportional reduction in monthly fees. After December 31, either party may terminate the agreement with 90 days’ notice.

These conditions suggest that delivery of the promised computing capacity will be closely watched. For a deal of this size, infrastructure readiness, GPU availability, energy supply and operational reliability will be critical.

The cloud agreement also comes as SpaceX pursues a proposed Terafab chip manufacturing and advanced computing project in Grimes County, Texas. Reuters reported that the project is central to SpaceX’s effort to expand beyond rockets and satellite communications into advanced computing infrastructure and domestic chip production.

The proposed Texas project has attracted both support and opposition. Supporters see it as a major investment that could strengthen US technology competitiveness, while local residents have raised concerns over water use, power demand, environmental impact and changes to the rural community.

The combination of major compute deals and large-scale infrastructure planning shows how SpaceX is building a wider technology ecosystem. The company is no longer only being valued on launch contracts or Starlink subscriptions, but also on its potential role in AI, semiconductors and next-generation computing.

For the wider AI industry, the Google-SpaceX agreement highlights how cloud capacity is becoming a strategic bottleneck. Even major technology companies with large existing infrastructure may need to lease external capacity to keep pace with customer demand.

This also explains why Nvidia GPUs remain at the centre of the AI race. High-performance GPUs are essential for many AI workloads, and access to large clusters of chips can influence how quickly companies can launch and scale AI products.

The agreement may also reshape how investors view SpaceX ahead of its IPO. Rather than being seen only as a space and satellite company, SpaceX is increasingly presenting itself as a multi-sector infrastructure company with exposure to aerospace, communications, AI computing and semiconductor manufacturing.

At the same time, the deal carries execution risks. Running large-scale AI infrastructure requires stable power, cooling, networking, data management and customer service capabilities. SpaceX will need to prove it can deliver these services at the standard expected by a major cloud customer such as Google.

For Google, the agreement could help ease short-term capacity pressure. However, it also shows how expensive the AI race has become. Paying nearly US$1 billion per month for compute capacity reflects the scale of investment needed to compete in enterprise AI.

The partnership also highlights the growing overlap between traditional cloud companies and non-traditional infrastructure providers. As AI demand expands, companies with access to chips, energy, data centre capacity or specialised infrastructure may become important suppliers even if they were not originally cloud providers.

Overall, SpaceX’s Google deal marks a major turning point in the company’s business story. The rocket company is now emerging as an AI infrastructure provider, using its computing assets to serve some of the world’s largest technology companies while strengthening its case to investors ahead of its expected IPO.

If SpaceX can successfully deliver the promised capacity, the agreement could become a powerful example of how the company can transform from a space-focused enterprise into a broader technology infrastructure giant.