SpaceX files for record-breaking IPO with rockets, AI, and Mars ambitions at the center

Skye Jacobs

Posts: 2,010   +58
Staff
Bottom line: SpaceX has filed for an IPO that could become the largest in history, giving investors a detailed look at a business that spans rockets, satellite networks, and AI infrastructure. The filing portrays a company growing quickly but spending heavily.

SpaceX reported $18.7 billion in revenue in 2025, up from $14 billion a year earlier, but posted a net loss of $4.94 billion after turning a profit in 2024. That trend has continued into 2026, with a $4.28 billion loss on $4.69 billion in revenue in the first quarter.

The results reflect heavy investment in AI and next-generation launch systems.

Much of that investment is tied to a push into computing. SpaceX plans to build orbital data infrastructure, including deploying 100 gigawatts of AI compute capacity each year using solar-powered satellites. The company describes this as part of a $28.5 trillion market, based on the idea that energy generation and computing can scale more efficiently in space than on Earth.

That effort accelerated following the company's acquisition of xAI. More than half of SpaceX's $20.74 billion in capital expenditures last year went toward AI-related initiatives. Losses tied to that segment climbed to $6.36 billion in 2025, up from $1.56 billion a year earlier. The company is also pursuing deals such as a potential $60 billion acquisition of AI startup Cursor, which includes significant financial penalties if terminated.

Despite the growing AI focus, SpaceX's revenue base remains heavily tied to Starlink. The satellite internet business accounted for roughly two-thirds of company revenue in the first quarter of 2026.

Despite the growing AI focus, SpaceX's revenue base remains heavily tied to Starlink. The satellite internet business accounted for roughly two-thirds of company revenue in the first quarter of 2026, fueled by subscriber growth from 2.3 million users in 2023 to 8.9 million in 2025. Revenue from Starlink reached $4.42 billion last year.

SpaceX is also developing upgraded satellites capable of connecting directly with smartphones, though those systems require more advanced launch capabilities than its current rockets can provide.

Credit: Bloomberg

That limitation makes Starship, the company's next-generation launch vehicle, more critical. Designed to be fully reusable and capable of carrying 100 to 150 tons into orbit, Starship is intended to lower launch costs and enable larger payloads, including the infrastructure needed for space-based computing. It is also central to longer-term goals, including missions to the Moon and Mars.

"The S-1 shows the launch business is loss-making, as we expected, due to heavy R&D spending. The filing points to an acceleration of Starship development which will increase the burn. Musk's goal of Starship reducing costs – to $185 a kilogram from a historical NASA average cost of $18,500 – is the key to the strategy of data centers in space and low Earth orbit satellite constellations," said George Ferguson, Bloomberg Intelligence senior defense analyst.

The IPO filing also details a governance structure that will keep control firmly in Elon Musk's hands. Through a dual-class share system, Musk holds 85.1% of the company's voting power.

Starship, however, is still in development and has faced multiple test failures, including several explosions in 2025.

The company notes that further delays or design changes could increase costs and require shifting resources across the business. Some next-generation Starlink satellites cannot be launched on current Falcon rockets, increasing reliance on Starship.

The IPO filing also details a governance structure that will keep control firmly in Elon Musk's hands. Through a dual-class share system, Musk holds 85.1% of the company's voting power, allowing him to maintain decision-making authority after the listing.

The filing outlines potential long-term incentives tied to milestones, including establishing a human settlement on Mars with at least one million inhabitants.

The offering, which could raise as much as $75 billion and value the company at up to $2 trillion, will hinge on whether investors are willing to back a business built on technologies that are still being developed. SpaceX is tying together launch, satellite connectivity, and large-scale computing, with each part depending on the others to scale.

Permalink to story:

 
It’s true that SpaceX is a business that’s still developing, but realistically its future business is modeled on profitable and unmatched EXISTING business. The article doesn’t mention the profit margins of Starlink service and Falcon 9 launches. It looks like that’s about 33% and 50% respectively. These are software-like margins, and the revenue for these are growing.

What the article doesn’t make clear is the vast majority or losses comes from AI. For example, in 2025 AI generated $3.2B in revenue and cost $9.6B to make it (with a net loss of $6.4B).

At the end of the day, SpaceX is a high profit + high reinvestment business.
 
Yet another StarSh!t launch scrub - https://aero-space.eu/2026/05/22/st...ch-attempt-of-its-next-generation-megarocket/

The "news" claims that this IPO will make Musk the world's first Trillionaire. Supposedly, that's predicated on 1,000,000 colonists living on Mars. I'm willing to bet that that target number for Mars Colonists will not be achieved in his lifetime.

Without Mars colonization, there's a limited market for space stuff, IMO. It would not surprise me if this IPO ends in a bursting bubble.

Keep going, fElon. Perhaps someday, reality will catch up with your BS.
 
It’s true that SpaceX is a business that’s still developing, but realistically its future business is modeled on profitable and unmatched EXISTING business. The article doesn’t mention the profit margins of Starlink service and Falcon 9 launches. It looks like that’s about 33% and 50% respectively. These are software-like margins, and the revenue for these are growing.

What the article doesn’t make clear is the vast majority or losses comes from AI. For example, in 2025 AI generated $3.2B in revenue and cost $9.6B to make it (with a net loss of $6.4B).

At the end of the day, SpaceX is a high profit + high reinvestment business.
Starlink is highly profitable.
SpaceX is reasonably profitable with upside potential once rocket development slows
xAI is a money blackhole BUT it potentially could be worth crazy money IF AI does great in the future.

Musk wants people to look at the profit of Starlink the revenue of Starlink + SpaceX and the growth potential of xAI when pricing the stock. Obviously that is not a wise way to price it but AI hype and Musk's ability to pitch his companies could just maybe get it across the IPO line.
 
They are already doing that, IMO.
Not nearly enough. Space X and Blue origin combined with the NASA budget is still a tiny fraction of AI spending. Space X got famous in no small part because space travel had stagnated for decades with chronic under funding.

AI funding globally was estimated to be about $300 billion
Spacetech was $6 billion in the same time frame.

Imagine where we would be if we were dumping $300 billion into spacetech for the last decade. Or nuclear fusion, or any other tech marvel that cna never get the funding that wunderprojects do.
 
"energy generation and computing can scale more efficiently in space"

Q1. How massive will a 9 GW solar panel have to be? (9 GW was the Mr. Wonderful article size)
Q2. How many meteor showers do we have each year?
 
Yet another StarSh!t launch scrub - https://aero-space.eu/2026/05/22/st...ch-attempt-of-its-next-generation-megarocket/

The "news" claims that this IPO will make Musk the world's first Trillionaire. Supposedly, that's predicated on 1,000,000 colonists living on Mars. I'm willing to bet that that target number for Mars Colonists will not be achieved in his lifetime.

Without Mars colonization, there's a limited market for space stuff, IMO. It would not surprise me if this IPO ends in a bursting bubble.

Keep going, fElon. Perhaps someday, reality will catch up with your BS.

The late Queen was worth some $150T on paper. Almost all of the Families are trillionaires, including the Rodhams, the Kennedys, the Stanleys, the Stewarts...
 
"energy generation and computing can scale more efficiently in space"

Q1. How massive will a 9 GW solar panel have to be? (9 GW was the Mr. Wonderful article size)
Q2. How many meteor showers do we have each year?
if it was for one installation, that would be 91 square kilometers on earth. In space solar panels are ~35% more capable then on earth, so maybe 70? However that 9 GW I dont think would be one satellite, it would be the collective of all orbiting AI sattelites.

After all, China has the most power hungry data center on earth, at 150MW of energy use, it takes up over 10 million square feet. No way you could build a satellite capable of holding 60 times that capacity, even before we get to cooling issues. That thing would be enormous.
 
Back