SpaceX Agrees to Buy Cursor Maker Anysphere

SpaceX said it has entered a formal agreement to acquire Anysphere, the company behind the AI coding assistant Cursor, in an all-stock transaction valued at roughly $60 billion. The announcement lands days after SpaceX completed one of the largest initial public offerings in market history, giving Elon Musk's rocket company a freshly minted public currency to spend on one of the most closely watched private software businesses in the industry.

The deal is expected to close during the third quarter, subject to customary conditions, according to a filing. Cursor will operate as a SpaceX subsidiary rather than being folded directly into the launch business, a structure that preserves the product's independent roadmap and its enterprise contracts while attaching it to a far larger balance sheet.

Anysphere had already been circled by SpaceX earlier in the year, when the two companies disclosed an option arrangement that gave the rocket maker a path to acquire Cursor at the same headline number. Converting that option into a signed agreement removes the ambiguity that had hung over Anysphere's fundraising conversations and its hiring pipeline for months.

Why a rocket company is buying a code editor

The strategic logic is less strange than the pairing first appears. Cursor sells software that writes, refactors, reviews and explains code, and it has become one of the fastest-growing developer products of the past two years, adopted broadly inside engineering organizations that previously standardized on general-purpose chat assistants. That places it in direct competition with coding tools shipped by Anthropic and OpenAI, both of which have made developer workflows a central pillar of their commercial strategy.

For SpaceX, the acquisition is an attempt to own a durable software franchise with recurring revenue rather than depend entirely on launch cadence, satellite broadband subscriptions and government contracts. It also gives the company a large internal customer for its own engineering: SpaceX writes an enormous amount of flight, ground and manufacturing software, and a code assistant tuned to that codebase is a plausible internal productivity lever as well as an external product.

The transaction is structured in stock, which matters. A cash deal of this size would have consumed a meaningful share of the IPO proceeds; paying in equity lets SpaceX conserve capital for its launch and satellite programs while transferring some of the valuation risk to Anysphere's existing shareholders, who become holders of a newly public security.

The price, and what it implies

Sixty billion dollars is an extraordinary figure for a company whose product shipped only a few years ago. It reflects a market that has decided AI-assisted software development is one of the few artificial intelligence categories with proven, repeatable enterprise willingness to pay. Coding is measurable: teams can point to review cycles, defect rates and time-to-merge, and the buyers making the decision are the same engineers using the tool.

It also reflects scarcity. There are only a handful of independent companies at Cursor's scale in the category, and each of the largest model labs has an incentive to own one rather than compete with all of them. Once SpaceX secured its option, the practical universe of alternative acquirers narrowed considerably.

Skeptics point to the obvious risk: Cursor sits on top of foundation models it does not own, and its economics are exposed to the pricing decisions of the labs that supply them. Anysphere has invested in its own models and in routing logic that reduces that dependence, but the structural question does not disappear with a change of ownership.

What it means for developers

For the millions of engineers who use Cursor daily, the near-term answer is likely continuity. Acquisitions of this size rarely produce immediate product change, and the subsidiary structure signals an intention to keep the team and roadmap intact through the close. The longer-term questions concern pricing, data handling and whether enterprise customers who compete with SpaceX in adjacent markets grow uneasy about their code assistant reporting to Musk.

Enterprise buyers have already begun asking those questions. Several large customers are expected to seek contractual assurances around data isolation and model training before renewing, a routine consequence of any change of control in developer infrastructure.

The wider consolidation wave

The deal caps a period in which the AI application layer has consolidated faster than almost anyone predicted. Model labs, cloud providers, and now an aerospace company have all concluded that owning the interface where work actually happens is more defensible than owning the model beneath it.

If the transaction closes on schedule, it will stand as one of the largest acquisitions of a private software company on record, and the clearest signal yet that the boundaries between hardware, aerospace and AI software companies have stopped meaning very much.