SpaceX Bets $10B on a $60B Option to Own Cursor — and Developer Data

What happened

SpaceX announced an agreement to acquire an option on Anysphere's Cursor — the AI-powered code editor that became the default IDE for a generation of developers who wanted Claude or GPT wired directly into their workflow. The deal structure is unusual: SpaceX pays approximately $10B upfront for a bundle of services and a call option to acquire Cursor outright at a $60B strike price. If Cursor's valuation exceeds $60B when the option matures, SpaceX gets a bargain. If it doesn't, they walk away with the services they already purchased — reportedly including access to developer interaction data and compute agreements.

This isn't an acquisition — it's a derivatives trade on the future of AI-assisted development, structured to give SpaceX asymmetric upside. The Reuters report confirms the option structure, while the NYT coverage frames it as part of Elon Musk's broader push to make xAI competitive in the coding model space, where Anthropic's Claude and OpenAI's Codex have established significant leads.

Anysphere, the company behind Cursor, had reportedly been fielding acquisition interest from multiple parties. Cursor's explosive growth through 2025 — driven by its tight integration of frontier models into the editing experience — made it one of the most valuable developer tool companies in history, despite being essentially a VS Code fork with a very good AI layer.

Why it matters

The $60B number sounds absurd until you understand what SpaceX is actually buying: not an IDE, but a dataset. Cursor sits between developers and their code all day, capturing the full context of how humans write, debug, refactor, and reason about software. That interaction data is arguably more valuable than any static code corpus for training the next generation of coding models.

As HN commenter nikcub pointed out, xAI has roughly 2GW of GPU capacity coming online this year with "largely not doing much outside of 'grok is this true.'" Combining that compute with Cursor's developer data creates a plausible path to a competitive coding model — something Grok has conspicuously lacked. The bull case writes itself: Cursor's data plus xAI's compute plus SpaceX's captive engineering workforce (thousands of developers building rockets and satellites) equals a vertically integrated coding AI pipeline.

The bear case is equally compelling. As one HN commenter (tombert) noted, they stopped using Cursor entirely once they figured out command-line Codex: "It just kind of seemed like a bloated, overpriced wrapper around what I could do with the included ChatGPT membership." This is the risk embedded in any $60B bet on a tool whose moat is primarily UX integration rather than proprietary model capability. Cursor doesn't own the models that make it valuable — it licenses them from Anthropic and OpenAI, the very companies it would be competing against under xAI's umbrella.

The naming discussion is revealing too. Commenter yungbeto flagged the obvious: both "X-Code" (Apple's Xcode) and "Codex" (OpenAI) are taken. It's a small point, but it signals a broader problem — xAI entering the coding tools market means competing with its current model suppliers. How long does Anthropic continue providing Claude API access to a Cursor that's now feeding data to Grok?

There's a third viewpoint worth considering: the "two sub-par models" thesis, articulated by anonymid. The idea is that xAI's Grok and Cursor's in-house Composer model are each individually behind the frontier, but combining their respective strengths — Grok's general reasoning plus Composer's code-specific fine-tuning — along with Cursor's proprietary data flywheel, could produce something competitive. This is the synthetic biology approach to AI development: combine two organisms that are individually unremarkable and hope the hybrid is greater than the sum of its parts. History suggests this works about 20% of the time in tech M&A.

What this means for your stack

If you're a Cursor user, the immediate question is data governance. Check your organization's Cursor subscription terms — the privacy policy around telemetry and interaction data is about to matter a lot more when the data recipient is xAI rather than Anysphere. Enterprise customers with strict data handling requirements should be reviewing this now, not after the option is exercised.

For teams evaluating AI coding tools, this deal reshuffles the competitive landscape. The major options now look like:

- GitHub Copilot — Microsoft/OpenAI stack, deeply integrated into VS Code and GitHub - Cursor — potentially xAI/Grok stack, future uncertain - Claude Code — Anthropic's CLI-first approach, no IDE lock-in - Augment, Cody, Continue — independent players that may look more attractive to teams wanting vendor neutrality

The option structure means this uncertainty persists for the duration of the option period. Cursor might remain independent. It might become an xAI product. That ambiguity alone will push some enterprise buyers toward alternatives with clearer roadmaps.

For the broader developer tools market, a $60B valuation anchor — even as an option strike — resets expectations for what AI coding tools are worth. Every AI code editor startup just got a new comparable for their fundraising deck, and every acquirer just got sticker shock. This is good for founders in the space and painful for anyone trying to acquire developer tooling at reasonable multiples.

Looking ahead

The most interesting variable isn't whether SpaceX exercises the option — it's how Anthropic and OpenAI respond. Cursor's entire value proposition depends on access to frontier models from companies that are now, effectively, its future competitor's rivals. If Claude or GPT-4 access gets restricted or degraded for Cursor, the product's appeal collapses regardless of how much GPU compute xAI throws at Grok. The $60B question isn't whether Cursor is worth that much today — it's whether it can maintain its model access long enough for xAI to build a competitive alternative. That's a race against the self-interest of its own suppliers, and it's not clear Cursor is favored to win it.