Why Musk’s $60 B Cursor Deal Is About More Than an Editor

What $60 B Means

On April 22, SpaceX’s official X account posted a tweet announcing a close collaboration between SpaceXAI and Cursor to build “the world’s best coding and knowledge‑work AI,” offering either a $60 billion outright acquisition or a $10 billion payment for compute resources.

Hours later, Cursor founder Michael Truell retweeted the announcement, expressing excitement about scaling Composer, and Cursor’s official account confirmed, “We’re partnering with SpaceX to improve Composer.”

Valuation Context

Cursor completed a D‑round financing in November last year at a $29.3 billion valuation; a year earlier (January 2025) it was valued at $2.5 billion, a 12‑fold increase. The $60 billion offer is more than double the latest valuation. For comparison, Microsoft’s 2018 acquisition of GitHub cost $7.5 billion, illustrating how AI is reshaping pricing in the developer‑tools market.

If the $60 billion deal closes, Cursor would become the highest‑valued AI programming tool, propelling the small Anysphere team into the top tier of Silicon Valley AI companies.

Two Paths Toward the Same Goal

The tweet reveals a staged collaboration rather than a simple binary choice. The first path grants Cursor access to SpaceX’s Colossus supercomputer for model training. Colossus, the xAI training cluster, is reported to have the compute power of one million Nvidia H100 GPUs. In exchange, Cursor would pay $10 billion for this compute.

The second path allows SpaceX to exercise an option later in the year to acquire Cursor outright for $60 billion, integrating the team, product, and users.

This design lets SpaceX evaluate how much Cursor’s models improve with Colossus; if the results are promising, the full acquisition follows, while the $10 billion fee serves as a compute‑rental cost if expectations aren’t met.

The Real Star: Composer

Cursor’s blog explains that the core of the deal is not the editor itself but the need for compute to train its proprietary model, Composer.

Composer, launched six months ago, is Cursor’s first‑generation intelligent coding model. It was upgraded to Composer 1.5 at the end of last year, expanding reinforcement‑learning scale by more than 20×, and this year introduced Composer 2 with continual pre‑training, claiming frontier‑model performance at lower cost.

Cursor’s team states, “We have ideas, data, and a team, but we’re bottlenecked by compute.” The missing piece is precisely the massive compute that Colossus provides.

Why Musk Is Betting on This

Understanding the deal requires looking at Musk’s broader strategy. In February, Musk deeply integrated xAI with SpaceX, combining xAI’s models and talent with SpaceX’s Colossus compute infrastructure.

Recent media reports note that xAI has begun renting compute to Cursor, and two former Cursor executives, Andrew Milich and Jason Ginsberg, quietly joined xAI. These signals indicate that collaboration started well before the public announcement.

Musk’s goal is to secure an “AI‑productivity gateway” that reaches professional software engineers—the most demanding and highest‑paying segment of the AI software market. Acquiring Cursor gives him direct access to this user base.

The partnership’s language, “distribution to expert software engineers,” is a strategic positioning aimed at investors.

Impact on Developers

In the short term, the user experience in Cursor will not change noticeably; the partnership is still an option, and the acquisition has not occurred. Composer 2 is already live, and developers can continue using it as before.

In the medium term, if Cursor’s self‑trained models improve, developers will benefit from better context understanding and multi‑file collaboration, potentially surpassing tools like Claude in efficiency gains.

In the long term, it is advisable not to rely on a single tool. Alternatives such as Claude Code and Codex should be monitored, as over‑reliance on one ecosystem can increase switching costs if the ecosystem shifts.