SpaceX filed an S-1 registration statement with the U.S. Securities and Exchange Commission on May 20, opening the books on a company that booked $18.7 billion in 2025 revenue and is now seeking to raise up to $80 billion at a $2 trillion valuation. If priced at the top of that range, the offering would be the largest in stock market history.
But the headline number obscures the more consequential story buried in the 277-page prospectus. The most valuable private company on the planet is no longer principally a rocket builder. After the February merger with xAI, SpaceX is presenting itself to public investors as a vertically integrated artificial intelligence platform that happens to own the world’s busiest launch business and its largest satellite network. Of the $28.5 trillion total addressable market the company claims, $26.5 trillion sits in AI. The rockets, in this telling, are infrastructure for something much larger.
An AI company that happens to launch rockets
The strategic repositioning is the part of the filing most likely to define how public investors price the stock. After absorbing xAI in February, SpaceX now frames itself as an artificial intelligence company with a uniquely useful orbital footprint.
The prospectus lays out a plan to launch orbital data centers starting in 2028, arguing that space offers structural advantages in cooling, solar power generation, and latency for global services. The company outlined in its filing its belief that it is uniquely positioned to deploy and operate data centers in orbit that can eventually achieve lower costs than terrestrial data centers over time due to its vertically integrated approach.
To feed those orbital servers, SpaceX disclosed a joint chip-fabrication project with Tesla, referred to internally as Terafab, designed to produce specialized AI silicon. Elon Musk’s compensation package is partly contingent on SpaceX deploying data centers capable of delivering 100 terawatts of computing power annually.
The market-sizing math is aggressive. SpaceX values the AI infrastructure opportunity at $2.4 trillion, enterprise AI applications at $22.7 trillion, broadband connectivity at $870 billion, and mobile at $740 billion. The company also lays claim to $760 billion in consumer AI subscriptions and $600 billion in digital advertising tied to its ownership of X and xAI products.
The numbers behind the trillion-dollar pitch
The financials underwrite the pivot in a specific way. SpaceX reported $18.7 billion in 2025 revenue, up 33% year-over-year, with $6.6 billion in adjusted EBITDA, according to SpaceNews’s review of the filing. Starlink, generating $11.4 billion of that total with subscribers doubling to 10.3 million in a year, is the cash engine paying for everything else. The launch business, still the public face of the brand, contributed $4.1 billion. Meanwhile capital expenditures nearly doubled to $20.7 billion in 2025, with $12.7 billion flowing into AI infrastructure versus $3.8 billion into rockets and $930 million into Starship in Q1 2026 alone — a spending mix that makes the AI story concrete. The price of that pivot shows up on the bottom line: a $4.9 billion net loss in 2025, reversing a $791 million profit in 2024, and another $4.3 billion lost in Q1 2026, according to Mint’s reading of the S-1.
What the filing leaves out
One absence is striking. SpaceX explicitly excludes China from its addressable market, both now and in the future, and instead flags the country as a competitive threat, Nikkei Asia reported. That is a sharp departure from Tesla’s playbook, where Chinese demand and manufacturing built much of the company’s scale.
The omission is more than rhetorical. Starlink does not operate in China. The Chinese government has accelerated its own megaconstellation programs, including Guowang and Qianfan, as direct competitors. SpaceX appears to be telling investors it has accepted, and priced in, a bifurcated orbital internet.
Control, governance, and the Musk question
Public investors will buy Class A shares carrying one vote each. Musk’s Class B shares carry 10 votes each. The filing confirms he holds more than 50% of shares outstanding and 85.1% of voting power.
His cash salary has not moved since 2019. It remains $54,080. The real compensation is structured as 15 tranches of roughly 66.7 million shares each, vesting against $500 billion valuation milestones up to $7.5 trillion, with an additional condition tied to establishing a permanent human colony of at least one million people on Mars. A separate 302 million-share grant, created when the xAI compensation plan was cancelled at the merger, vests against the 100-terawatt data-center milestone.
The board, disclosed publicly for the first time, includes President and COO Gwynne Shotwell, CFO Bret Johnsen, longtime backers Steve Jurvetson, Luke Nosek, Randy Glein, Ira Ehrenpreis and Antonio Gracias, plus Google executive Donald Harrison. Observers note that Musk maintains significant control over SpaceX’s governance.
What it means for the broader market
Goldman Sachs is leading the offering. The company is targeting a Nasdaq listing under the ticker SPCX, with a roadshow as early as June 4 and pricing as soon as June 11. Class A shares will trade on both the main Nasdaq exchange and the recently launched Nasdaq Texas venue, fitting for a company headquartered in Starbase.
A successful debut would unlock liquidity for early backers like Peter Thiel’s Founders Fund, which invested in 2008, and Valor Equity Partners, whose founder Antonio Gracias is the second-largest shareholder with 7.3% of common stock. It would also reopen the door for other capital-hungry private giants. Anthropic and OpenAI are watching closely.
Retail demand has already started to price the company in unofficial venues. A synthetic perpetual futures contract tied to SpaceX equity launched on the Hyperliquid decentralized exchange on May 18, opening at a reference price of $150 and spiking as high as $216 on its first day before settling at $202.89 with $33 million in volume. That implied valuation pushes north of $2.4 trillion.
The risk profile public investors are being asked to accept
Strip away the rhetoric and the SpaceX prospectus presents a recognizable shape: a cash-generating connectivity business funding two enormous bets that may not pay off for years. Starship is still in test-flight cadence. Orbital data centers do not yet exist. The xAI integration is months old. Net losses are accelerating, not narrowing.
The bet investors are being offered is on execution at a scale no company has attempted before, governed by a founder with near-total voting control, in markets the company itself has had to invent to justify the valuation. The S-1 describes the company as having capabilities in identifying, developing, and commercializing new multi-trillion-dollar markets that did not previously exist.
That framing is what makes this IPO different from any space-sector listing that has come before it. Investors who buy SPCX at a $2 trillion valuation are not pricing rockets, satellites, or even Starlink subscribers — those businesses, generously valued, do not get close to the number. They are pricing the orbital data center thesis, the Terafab silicon roadmap, and the xAI software stack that has to ride on top. The space industry’s smaller listed players offer a cautionary counterpoint: Spire Global is currently navigating NYSE noncompliance proceedings over a delayed annual report, a reminder that quarterly discipline tends to grind down even visionary stories. For competitors — from legacy launch providers to AI hyperscalers like Anthropic and OpenAI — a successful SPCX debut would validate orbital compute as a fundable category and accelerate the race to build it. For everyone else, the question is whether the public market will underwrite a company that has explicitly told it that rockets are no longer the point.
