SpaceX Dangles $60B Buyout Option in Massive Cursor AI Deal

TL;DR

• SpaceX partners with AI coding startup Cursor in a $10 billion collaboration using the Colossus supercomputer.
• Deal includes an option for SpaceX to acquire Cursor outright for $60 billion.
• Move follows SpaceX’s xAI integration and positions the rocket company deeper into AI infrastructure ahead of a rumored IPO.
• Partnership aims to build what SpaceX calls ‘the world’s most useful models’ by combining Cursor’s coding tools with massive compute power.

SpaceX Writes a $10B Check to an AI Coding Startup

SpaceX announced a partnership with Cursor, an AI-powered coding assistant, that includes a $10 billion collaboration centered on the company’s Colossus AI training supercomputer. The deal also grants SpaceX an option to acquire Cursor entirely for $60 billion. The partnership was reported by The Straits Times, citing the companies’ joint announcement.

According to the announcement, combining Cursor’s software and product expertise with SpaceX’s Colossus AI training supercomputer will enable the company “to build the world’s most useful models.” That’s a bold claim in a market already crowded with coding assistants — GitHub Copilot, Replit, Codeium, and a dozen others all promise to make developers faster.

But SpaceX isn’t playing the same game as those competitors. This isn’t about selling subscriptions to developers. It’s about locking down the infrastructure and talent to train foundation models at a scale most AI labs can only dream about.

Why SpaceX Is Suddenly an AI Kingmaker

Let’s be clear: SpaceX is a rocket company. Or it was. Now it’s also sitting on one of the most powerful AI training clusters on the planet, and it’s using that leverage to pull promising startups into its orbit — literally and figuratively.

The $10 billion collaboration gives Cursor access to compute resources that would cost a fortune to rent from AWS or Google Cloud. For a startup, that’s transformative. It means training runs that would take weeks can happen in days. Models that would blow through a Series C can get built on someone else’s dime.

And SpaceX? It gets first crack at whatever Cursor builds. If the models pan out, SpaceX can exercise that $60 billion option and own the whole stack. If they don’t, SpaceX still gets cutting-edge coding tools and deeper hooks into the AI talent market. It’s a bet with a floor and no ceiling.

This follows SpaceX’s integration of xAI, Elon Musk’s other AI venture, which already gave the company a toehold in large language model development. By adding Cursor to the mix, SpaceX is building a vertically integrated AI operation that spans infrastructure, tooling, and application layers. That’s not a side project. That’s a pivot.

I can’t help but think this is less about SpaceX needing better code and more about Musk positioning his companies as the compute backbone for the next wave of AI startups. Colossus isn’t just training rockets anymore — it’s training the tools that train everything else. And if you control the foundry, you control who gets to build.

Cursor Gets Compute, SpaceX Gets Optionality

For Cursor, this deal is a lifeline and a gamble. Access to Colossus means the startup can compete with OpenAI and Anthropic on model quality without raising another mega-round. But that $60 billion acquisition option hangs over everything. If SpaceX decides to pull the trigger, Cursor’s independence evaporates.

The structure here is unusual. Most partnerships don’t come with a preset buyout price. That $60 billion figure suggests SpaceX and Cursor’s investors have already negotiated the endgame. Either Cursor hits certain milestones and SpaceX buys it, or the partnership dissolves and both sides walk away. There’s no middle ground.

For context, GitHub sold to Microsoft for $7.5 billion in 2018, and Copilot has since become one of the most widely used AI coding tools in the world. If Cursor can capture even a fraction of that market — and train models that generalize beyond code — a $60 billion valuation starts to look reasonable. But that’s a big if.

The competitive stakes are real. GitHub Copilot is already embedded in millions of developers’ workflows. Replit is building an entire cloud IDE around AI pair programming. Cursor needs to leapfrog both in capabilities, and it needs to do it before the window closes. SpaceX’s compute gives it a shot. Whether it’s enough depends on execution.

SpaceX’s AI Strategy Comes Into Focus Ahead of IPO Rumors

This partnership is part of a broader pattern. After absorbing xAI, SpaceX has been quietly positioning itself as more than a launch provider. It’s becoming an AI infrastructure play, and that shift matters a lot if the company is serious about going public.

Investors love optionality. A rocket company with a monopoly on reusable boosters is valuable. A rocket company with a monopoly on reusable boosters and a controlling stake in next-gen AI tooling? That’s a different valuation entirely.

The timing also aligns with broader trends in AI development. Training costs are skyrocketing, and access to cutting-edge compute is becoming the bottleneck for startups. SpaceX can rent out Colossus capacity, offer partnerships like this one with Cursor, or even spin out an AI-as-a-service business. Each path unlocks new revenue streams that have nothing to do with launching satellites.

And let’s not ignore the talent angle. By partnering with Cursor, SpaceX gets a front-row seat to one of the most promising teams in AI tooling. If the collaboration works, those engineers could end up inside SpaceX permanently. If it doesn’t, SpaceX still learned what not to build. It’s a recruiting funnel disguised as a partnership.

What Happens If SpaceX Pulls the Trigger on That $60B Option

If SpaceX exercises the acquisition option, it would mark one of the largest AI deals ever — dwarfing Google’s DeepMind acquisition and rivaling Microsoft’s OpenAI investment in scale. That would signal SpaceX is serious about competing in the foundation model race, not just providing infrastructure for others.

But acquisitions this size come with integration risk. Cursor’s team would need to mesh with SpaceX’s engineering culture, which is famously intense. The models would need to prove their value beyond coding — SpaceX’s language about “the world’s most useful models” suggests ambitions that stretch into robotics, logistics, and autonomous systems. If Cursor can’t deliver on that vision, $60 billion starts to look like a very expensive acqui-hire.

Watch how Cursor’s product evolves over the next 12 months. If the startup starts shipping models that handle not just code but natural language reasoning, multimodal tasks, or real-time decision-making, that’s a sign SpaceX sees broader applications. If Cursor stays narrowly focused on developer tools, the acquisition becomes less likely and the partnership becomes more transactional.

Also watch for other SpaceX partnerships with AI startups. If this is the first of several deals — compute for equity, with acquisition options baked in — then SpaceX is building a portfolio strategy. That would make Colossus the centerpiece of a new kind of AI venture fund, one where the currency is training runs instead of cash.