Summary of Key Points
Elon Musk's SpaceX acquired the AI programming tool Cursor in a all-stock transaction worth $60 billion. This deal not only caused SpaceX's stock price to soar ( momentarily surpassing that of Microsoft and Amazon) but also exposed a fatal weakness in AI application companies: their reliance on underlying models, which can become a bottleneck. Cursor, a “star application” with annual revenue exceeding $2 billion over three years, was ultimately acquired due to its lack of an independent core model and the competition from Anthropic, a company that develops such models. This case highlights the brutal trend in the AI industry where “models are devouring applications”—underlying model companies are bypassing the intermediate application layer to directly capture the market.
I. The $60 Billion All-Stock Acquisition: Transaction Details and SpaceX’s Stock Price Boom
This acquisition was not a spur-of-the-moment decision but had been planned in advance:
- Transaction Method: A complete exchange of stocks, with Cursor’s shareholders receiving SpaceX shares instead of cash. The valuation of $60 billion (approximately 406.2 billion RMB) represents a 20% premium over Cursor’s previously planned fundraising target of $50 billion.
- Option Arrangement: In April of this year, SpaceX obtained the right to acquire Cursor first. However, due to the postponement of its IPO, both parties agreed that they would pay $10 billion regardless of whether the acquisition went through. If the deal went through, Cursor would be integrated into SpaceX’s portfolio; if not, SpaceX would still pay $10 billion as a compensation for their cooperation efforts.
- Stock Price Reaction: Upon the announcement, SpaceX’s stock price rose by 16% during trading, briefly exceeding Microsoft’s market value of $2.94 trillion. By the end of the day, it surpassed Amazon to become the fourth-largest company in the United States, with its stock price up nearly 50% since the IPO.
In short, Musk used SpaceX’s shares to acquire Cursor, simultaneously boosting his own company’s stock price.
II. Why Does Cursor Worth $60 Billion? The Myth of Growing from Zero to $2 Billion in Three Years
Cursor is considered a “money-making machine” within the AI industry, with several notable features:
- Growth Rate: Founded in 2022 by four MIT graduates in their 20s, the company’s annual revenue grew from zero to $100 million in January 2025, to $500 million in June, and to $2 billion in February 2026—breaking records in the B2B software industry within just three years.
- Profitability: It has a high conversion rate of paid users, with each employee generating more revenue than many prominent Silicon Valley companies (for example, a team of 100 people earning $2 billion annually, averaging $20 million per person).
- Capital Recognition: The company received investment from OpenAI during its seed round and was valued at $50 billion before the acquisition. Musk’s offer of $60 billion underscores its potential value.
So, why was such a profitable company acquired? Because it has a critical flaw.
III. A Fatal Weakness: Dependence on Others’ Models
Cursor’s business model relies on third-party models:
- Core Dependency: It does not have its own AI model and charges users a subscription fee to use the API for Anthropic’s Claude model. At its peak, Cursor contributed 40%-50% of Anthropic’s revenue, making it one of its top sales channels.
- Reversal of Fortune: In 2025, when Anthropic launched its own programming tool, Claude Code, its user base surpassed that of Cursor within half a year. Anthropic later tightened access to the API, cutting off Cursor’s revenue stream. Developers complained that Cursor’s data optimization efforts reduced efficiency compared to Claude Code.
- Failed Self-Saving Efforts: Cursor attempted to develop its own model, Composer 2, claiming it outperformed Claude Code, but it was revealed to use Musk’s Kimi model, indicating that it remained dependent on others.
In essence, Cursor is like a car company that leases engines from others; once the engine manufacturer starts manufacturing cars itself, Cursor can no longer survive on its own.
IV. “Models Devouring Applications”: The Cruel Reality of the AI Industry
This acquisition has punctured the bubble in the AI application layer:
- What does “models devouring applications” mean? Underlying model companies (such as Anthropic and OpenAI) originally sold their models to application developers, but now they are creating their own applications to directly compete with them. For example, Anthropic’s Claude Code serves developers without going through Cursor.
- Impact on Application Companies: Traditional software companies charge based on the number of accounts, while AI model companies charge per usage (e.g., a fee for processing a certain amount of data). Model companies have a technological advantage that allows them to reach users directly, eroding the protective barriers of application companies.
- Industry Impact: The global software market is worth nearly $500 billion, and this trend may lead to a redivision of the market by model companies. Most application companies, unlike Cursor, will not be able to fetch a price tag of $60 billion for their business and may either be acquired or go out of business.
This is akin to running an internet-famous bubble tea shop using someone else’s exclusive recipe. When the recipe company opens its own chain stores, it not only takes away your customers but also raises prices—either you get acquired or you have to close down.
V. Musk’s Strategy: Addressing AI Weaknesses and Building an “AI + Aerospace” Ecosystem
Why did Musk spend $60 billion on Cursor?
- Strengthening Weaknesses: He has publicly acknowledged that his xAI technology lags behind in programming tools. Acquiring Cursor provides him with a mature product and its user base of professional engineers.
- Resource Integration: SpaceX possesses supercomputers with millions of H100 cores, while Cursor has programming tool capabilities and distribution channels. Together, they can create the “most practical AI models” to enhance competitiveness in software development.
- Expansion of the Ecosystem: Musk’s business empire already includes SpaceX (aerospace), Tesla (automobiles), and xAI. With the acquisition of Cursor, his “AI + aerospace” strategy becomes clearer: using AI to optimize aerospace software or leveraging aerospace computing power to advance AI development.
For Musk, this is more than just acquiring a tool; it’s a crucial step in completing his AI ecosystem.
Conclusion
Cursor’s fate serves as a reminder to all AI application companies: without their own core models, even if they are profitable, they are essentially working for model companies. In the future, either they must develop their own models or rely on larger entities (like Cursor, which was acquired). Otherwise, they will struggle to survive in the wave of “models devouring applications.” Musk’s investment in Cursor is another major move in his bid to gain an advantage in the AI competition by combining computing power with application capabilities.