SpaceX is paying $60B for a VS Code fork that wraps rivals' models

A rocket company bought a code editor with stock printed last week

SpaceX has agreed to acquire Anysphere, the maker of Cursor, at a $60 billion valuation, a deal Reuters reported and SEC filings linked in the thread confirm. The structure is the story: this was set up months ago as an option locked in at this same valuation, with a $10B breakup fee if SpaceX walked. After SpaceX raised on the public market, it exercised.

The number that detonated the thread came straight from the IPO pitch. SpaceX told investors it sees an addressable market for AI products worth $26 trillion: roughly the entire annual GDP of the United States.

"SpaceX told investors during the IPO process that it sees an addressable market for AI products worth $26 trillion, roughly equivalent to U.S. GDP." This is unhinged. — the thread's top-karma commenter

What you actually get for $60B: a fork, a harness, and Kimi K2.5

Strip the valuation and Cursor is a VS Code fork wrapping other people's models, with one home-grown model, Composer 2.5, based on the open Kimi K2.5. So why would a rocket company want it? The thread's best guesses converge on data and distribution, not technology: usage data to train Grok's coding, plus a ready-made customer base to fight Claude Code and Codex on the mid-market.

Cursor is an extension for VS Code, a harness and a bunch of prompts… What are they after for here? Customer base? Talent? — a developer building a spaced-repetition knowledge graph

Not everyone agrees Cursor is dead weight. One user reports Composer 2.5 already ships as the default in Grok Build.

Compose 2.5 is the default model in Grok Build. And it's quite incredible. It's comparable to Opus 4.7 but faster and incredibly cheaper. — a founder building a Toulouse events app

"Definitely not a bubble"

The sarcasm ran thick. One commenter swapped "AI space" for "tulip space" to describe the wealth-per-user; another imagined a future documentary host asking the audience "Where were the signs?" Even tptacek, dampening the drama, noted the deal was effectively public months ago at this price: the news is that it became binding, not that it's new.

The sharpest tension isn't optimist versus pessimist. It's that the skeptics concede they're losing.

$60b is genuinely insane… for a product with arguably no defensible moat. Congrats to the Cursor team though. — a commenter tracking his own P(doom) at 98%

The asterisk: nobody is paying cash

The instinct that SpaceX torched 80% of a fresh raise to buy an IDE is the wrong read. It pays in equity — "monopoly money," as one commenter put it — so the cost is dilution, not a drained treasury. That reframes the bet: SpaceX is converting last week's public-market enthusiasm into a coding-data pipeline for Grok, at a price it pre-negotiated before the run-up. Whether the $26T pitch survives contact with reality is a separate question. The deal itself is a masterclass in buying with paper while it's still expensive.