News broke yesterday via TechCrunch that is bigger than any model release: SpaceX is not just a partner of Cursor — it holds an option to acquire Cursor for $60 billion.
This is not a rumor. SpaceX announced it itself.
Breaking down the structure
SpaceX and Cursor signed a joint agreement aimed at "developing the next generation of AI for programming and knowledge work," backed by SpaceX's Colossus supercomputing cluster operating in Memphis — claimed to be equivalent to 1 million Nvidia H100 GPUs in compute power.
Within this cooperation framework, SpaceX holds two options:
- Option A: Acquire Cursor directly for $60 billion
- Option B: Buy the technical results of this collaboration for $10 billion
The $60 billion figure is not arbitrary. Cursor is currently negotiating a funding round at a $50 billion valuation. SpaceX's option price is 20% above that market valuation. That premium signals that Musk believes the company is worth more than its current price tag.
Then there is xAI's role: Musk's other company, xAI, is leasing chip compute to Cursor for model training — meaning part of Cursor's model training infrastructure already runs on Musk's hardware.
And two names: Cursor core engineers Andrew Milich and Jason Ginsberg recently moved to xAI, reporting directly to Musk.
Leasing you compute, poaching your engineers, holding an acquisition option — this is not just a strategic partnership. It is an encirclement.
Why Cursor is worth this much
Why would SpaceX be willing to pay $60 billion for Cursor? The numbers tell the story:
| Milestone | Valuation / ARR |
|---|---|
| January 2025 | $2.5 billion valuation |
| May 2025 | $9 billion valuation |
| November 2025 | $29.3 billion valuation ($2.3B Series D) |
| February 2026 | ARR surpasses $2 billion |
| April 2026 | Negotiating valuation of $50 billion |
| End of 2026 (projected) | ARR exceeds $6 billion |
In three years, ARR went from zero to $2 billion. Now Cursor is raising another $2 billion in this round, led by a16z and Thrive, with Nvidia also participating and Battery Ventures as a new investor.
Cursor also made a smart cost-control move: it switched some AI calls from Anthropic/OpenAI to the cheaper Kimi (Moonshot AI). Enterprise accounts are already gross-margin positive, while individual developer accounts are still loss-making.
The real meaning of this chess move
Cursor's direct competitors are Anthropic's Claude Code and OpenAI's Codex.
If Musk ultimately acquires Cursor, the chain is complete:
- SpaceX has Colossus compute
- xAI has the Grok model
- Cursor has the product and a $2 billion ARR user base
- Two core engineers are already inside
A possible future path: replace Cursor's current underlying inference models (Claude/GPT) with Grok, replace rented Nvidia compute with xAI chips, and turn Cursor from a "neutral AI IDE" into part of the Musk ecosystem.
The impact on Anthropic would be significant. Cursor is currently the entry point for a large portion of developers using Claude. If that entry point falls under Musk's control, it effectively cuts off one of Claude's core distribution channels in the developer market.
Of course, an option is not an execution. Whether it will be exercised won't be known until "later in 2026." Cursor's founding team has not made a public statement, and a16z has not responded.
But Musk's moves in AI coding tools are no longer "possible involvement" — he is playing a very concrete game.
Sources: SpaceX is working with Cursor and has an option to buy the startup for $60B (TechCrunch); CocoLoop, Sources: Cursor in talks to raise $2B+ at $50B valuation as enterprise growth surges (TechCrunch); AI startup Cursor in talks to raise $2 billion funding round at valuation of over $50 billion (CNBC); Cursor In Talks to Raise $2 Billion at Over $50 Billion Value (Bloomberg)