📊 Full opportunity report: SpaceX Owns Every Layer of AI Now. The Model Is Still the Weak Link. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
SpaceX has purchased Cursor, a profitable AI coding company, for $60 billion, giving it control over all AI layers from hardware to applications. The company’s AI models remain its weak point, despite its integrated infrastructure.
SpaceX has completed a $60 billion all-stock acquisition of Cursor, a profitable AI coding company, thereby controlling every layer of the AI infrastructure — from hardware and data centers to applications and models. This move consolidates SpaceX’s position as a dominant, vertically integrated AI powerhouse, though its AI models are still considered its weakest link, according to industry analysts.
On June 16, 2026, SpaceX announced it exercised its option to acquire Cursor, a company founded in 2022 that specializes in AI coding tools, for $60 billion in all-stock. The deal is expected to close in the third quarter of 2026, after which Cursor will become a wholly owned subsidiary of SpaceX. Prior to the acquisition, Cursor achieved approximately $4 billion in annual revenue, with notable clients including major AI labs such as Anthropic and Google, which lease significant compute resources from Cursor’s infrastructure.
SpaceX’s acquisition grants it control over the entire AI stack: from its supercomputers, like the Colossus clusters in Memphis, to its proprietary silicon, research labs, and application layer through Cursor and its Grok models. This vertical integration makes SpaceX unique among Western tech firms, positioning it as a fully integrated AI conglomerate. The company’s compute capacity includes roughly 555,000 Nvidia GPUs, with plans to deploy satellites as orbital data centers, and on-site gas generators to power its infrastructure.
Despite owning the hardware and applications, SpaceX’s AI models, such as the Grok line and Cursor’s latest models, are still considered underperforming relative to industry standards. Industry insiders point out that the models’ FLOP utilization remains low, and training efficiency has been a challenge, partly due to hardware architecture mismatches and low parallelization efficiency. Elon Musk has publicly acknowledged these issues, emphasizing that the models are still in development and that the infrastructure is the strongest aspect of the company’s AI ambitions.
SpaceX owns every layer
of AI now
The $60B Cursor buy completes the stack: power, compute, research, model, app, distribution. But owning every layer isn’t winning every layer — and the model is the weak one.
(Anysphere)
You can buy a coding app and a model team. You can’t buy the research lead that makes your foundation model the one everyone else builds on — which is why Anthropic pays Musk $1.25B/month, not the other way around. Owning every layer bought SpaceX the right to attempt the hard thing. It hasn’t done it yet.
Implications of SpaceX’s Complete AI Control
This acquisition positions SpaceX as the most vertically integrated AI company in the West, controlling hardware, data centers, research, and applications. It consolidates a significant portion of the AI infrastructure industry, with major clients like Anthropic and Google leasing compute power, thus shaping the competitive landscape. However, the ongoing challenge remains: despite owning the entire stack, SpaceX’s AI models are not yet optimized for production, which could limit its immediate competitive edge. The move signals a strategic bet on infrastructure dominance, but the success depends on model performance improvements and operational efficiencies.
Nvidia GPUs for AI training
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background on SpaceX’s AI Infrastructure Expansion
Over recent years, SpaceX has invested heavily in AI infrastructure, building the Colossus supercomputers capable of training massive models at unprecedented speeds. The Memphis-based clusters, which now utilize approximately 555,000 GPUs, were built rapidly—initially in 122 days for 100,000 GPUs—and cost billions of dollars. The company has also developed proprietary silicon and is exploring orbital data centers powered by solar satellites. Meanwhile, SpaceX’s AI research arm, xAI, was integrated into the company in February 2026, with the Grok model line serving as its foundation.
Prior to the Cursor acquisition, other industry giants like OpenAI and Microsoft had approached Cursor, but the company prioritized independence, fueling speculation about a strategic move by SpaceX. The company’s control over compute resources is exemplified by leasing its hardware to rivals like Anthropic and Google, both of which pay hundreds of millions annually for access, due to low utilization of its clusters. This leasing strategy has created a unique dynamic where SpaceX’s infrastructure is both a competitive asset and a revenue generator.
“We’re comfortable leasing our compute because training has moved on, but we reserve the right to reclaim it if necessary.”
— Elon Musk
NVIDIA DGX Spark™ – Personal AI Desktop Supercomputer – Desktop GB10 Grace Blackwell Chip
Supercomputer performance directly to your desk in a compact, energy-efficient design, enabling enterprise-scale AI and high-performance computing right…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Unresolved Challenges in Model Performance
While SpaceX now owns the entire AI stack, its models, including Grok and Cursor’s latest offerings, are still considered underperforming relative to industry benchmarks. Industry insiders cite low FLOP utilization and inefficient parallelization as key issues, and it is unclear how quickly these problems will be resolved or how they will impact the company’s competitive stance in AI development.
AI Systems Performance Engineering: Optimizing Model Training and Inference Workloads with GPUs, CUDA, and PyTorch
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Future Developments and Model Optimization Efforts
SpaceX is expected to focus on improving its AI models’ efficiency and performance over the coming months, leveraging its integrated infrastructure. The company may also expand its orbital data center initiatives and further develop its proprietary silicon and hardware capabilities. The closing of the Cursor deal in Q3 2026 will mark a significant milestone, after which the company will likely announce new models and operational strategies to capitalize on its full-stack control.
enterprise AI server racks
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Why did SpaceX acquire Cursor for $60 billion?
SpaceX aimed to control every layer of the AI stack, from hardware to applications, and acquired Cursor to gain a profitable, established AI coding product and a team capable of developing advanced models, integrating them directly into its infrastructure.
What are the main challenges facing SpaceX’s AI models?
The models currently underperform in terms of FLOP utilization and parallelization efficiency, limiting their readiness for production use. Improving these aspects is a key focus for the company’s AI development efforts.
How does owning all AI layers give SpaceX an advantage?
Vertical integration allows SpaceX to optimize hardware, training, and deployment processes, reduce costs, and potentially accelerate AI development, although model performance remains a critical factor for success.
Will SpaceX continue leasing its compute resources to rivals?
Yes, leasing its infrastructure is profitable and strategically advantageous, but the company retains the option to reclaim compute resources if needed, especially as it works to improve its own models.
What is the significance of orbital data centers in SpaceX’s AI plans?
SpaceX’s orbital data centers aim to provide scalable, solar-powered compute capacity in space, potentially revolutionizing AI infrastructure by enabling data processing at a planetary or even orbital level.
Source: ThorstenMeyerAI.com
