SpaceX’s $2 trillion reality check meets hard limits

SpaceX’s $2 – SpaceX is pitching an IPO that could value it at up to $2 trillion, but the road to that dream runs through bottlenecks it can’t ignore: delayed rockets, a massive launch backlog, and an AI bet still burning cash. The company’s biggest strength—Starlink—sits a

By the time investors gather around SpaceX’s IPO prospectus. the company’s pitch is already clear: a future in space that’s big enough to justify a $1.75 to $2 trillion valuation. SpaceX wants to raise about $75 billion—while selling the idea that it’s building infrastructure for trillion-dollar markets on the Moon. Mars. and beyond.

But the numbers and timelines inside that same prospectus—and the delays across the broader launch market—make one thing hard to look away from. SpaceX’s next chapter. especially its Starship-dependent ambitions. runs into a blunt reality: getting payloads to orbit at the needed pace remains the chokepoint.

SpaceX completed 165 orbital launches in 2025, accounting for nearly 51% of total launches worldwide and sending aloft 85% of all satellites last year. Two companies outside of China and Russia—SpaceX and Rocket Lab—are delivering payloads into orbit frequently and reliably. Rocket Lab recorded 21 successful launches in 2025, including orbital and suborbital missions.

The scale gap is just as unforgiving as the schedule gap. SpaceX’s Falcon 9 can carry a maximum capacity of 13,000 kilograms, while Rocket Lab’s Electron maxes out at about 300 kilograms.

SpaceX’s bottleneck is not capacity in the past. It’s capacity in the future.

Starship, the rocket central to SpaceX’s plan, was developed at a cost of more than $15 billion. In its current fully reusable configuration. Starship is designed to deliver 100 metric tons to Earth orbit—and double that in future versions. In the company’s own words, Starship is central to unlocking growth through SpaceX’s “unique vertically integrated business model.”.

Starship has made 12 R&D flights since 2023, with six successful. The company’s most recent launch last week was mostly successful. It expects to start payload delivery to orbit in the second half of 2026.

That timeline matters because competitors are also behind. Rocket Lab’s medium-lift Neutron was originally expected in 2024, but the inaugural launch has been pushed back to the end of 2026 at the earliest after a first-stage propellant tank ruptured in January.

In the U.S., the alternative launch providers—United Launch Alliance and Blue Origin—are also struggling. ULA’s Vulcan Centaur. the heavy-lift successor to its 20-year-old Atlas V. has been effectively grounded after three flights due to debris shedding from a booster during a February launch. Blue Origin’s New Glenn put one NASA payload into orbit last year. but in April it missed the orbital target for a satellite for AST SpaceMobile. and the misplaced satellite had to be de-orbited.

The delays have turned into a backlog with consequences that show up in real business deadlines.

Amazon is nearing a July 30 deadline with the FCC to have half of its proposed LEO communications satellite constellation deployed. Amazon has contracted 38 Vulcan launches and 12 with New Glenn, plus 18 through Airbus subsidiary Arianespace and three with SpaceX. So far, it has deployed about 10% of its planned 3,232-satellite constellation. In its appeal of the FCC deadline, the company points to “a near-term shortage in launch capacity.”.

“The age of AI and abundance of capital and all the rest of it, [launch] is still freaking hard,” Rocket Lab founder and CEO Peter Beck said. “It’s really rare to get all the things to come together to make it work. There’s no partial success. It’s success or fail.”

SpaceX’s defenders argue the market will eventually add players. James Bruegger, chief investment officer at space-focused Seraphim VC, says additional competitors will emerge. The market. he argues. has “never been one-size-fits-all.” As more rockets of different sizes. classes. and capabilities come online. SpaceX “isn’t going to have the market all to itself.”.

Still, any new entrant faces a steep challenge: Starship’s price ambition. SpaceX’s IPO prospectus states that fully loaded Starships could eventually reduce the historical average launch cost per kilogram of $18,500 by 99% or more.

Even if that comes true, the timing problem remains.

SpaceX’s AI pitch leans even harder on overcoming the same constraint.

To justify a valuation approaching $2 trillion, SpaceX is not selling investors only rockets or broadband. It is also pushing space-based AI infrastructure—massive data centers orbiting Earth. For now, AI looks like the company’s Achilles’ heel.

SpaceX’s xAI division, acquired in February, ran an operating loss of $6.35 billion in 2025 while consuming 61% of total capital expenditures. In April, SpaceX secured an option to buy Cursor, the AI coding startup, for $60 billion, and layoffs and staff departures followed.

Musk is betting that fusing SpaceX’s fortunes with AI will compound the opportunity. SpaceX’s prospectus forecasts a total addressable market of $26.5 trillion for its AI initiatives. In the shorter term. it is hedging through a deal that could bring in $1.25 billion a month to xAI through May 2029 by selling unused compute capacity to Anthropic.

There is already plenty of data being collected in space—for intelligence, surveillance, reconnaissance, and other purposes. But the company argues it doesn’t have the capacity right now to beam it all back to Earth. “If you could process it [in space] and just send down the interesting insights. that is the near-term capability to lean into. ” said Taylor Sargent. a partner at Industrious VC.

Companies have begun deploying on-orbit edge computing nodes, but they’re running into launch bottlenecks as they try to get their assets into space. SpaceX suggests that scenario is where its own scale can matter most.

Building large-scale AI data centers in space is a different proposition. Proponents argue that as it gets harder to find land. water. and energy for new AI data centers on Earth. space becomes more appealing because of abundant solar power and no neighbors. SpaceX’s prospectus claims space is “the only truly scalable solution to the challenge of accelerating demand for compute relative to terrestrial energy constraints.”.

The target is enormous enough to feel unreal on paper: SpaceX hopes to deploy 100 gigawatts of computing capacity in space each year. The prospectus says the effort “will require thousands of launches per year” and the transport of approximately one million metric tons to orbit annually. SpaceX filed plans with the FCC in January for an “orbital data center” with up to 1 million satellites to power advanced AI models and applications.

For those structures to exist, SpaceX and its competitors will need exponentially more launch capacity. They will also need to build in space—using robots and automation, presumably—and solve operational issues such as heat dissipation and radiation hardening.

Still, investors are leaning bullish. “These are all solvable problems, in our view,” Bruegger said. “I have a high degree of confidence that a time will come where a meaningful amount of AI compute is being done in orbit. by the back end of this decade. The step change will be these mega launch vehicles flying regularly.”.

Beck offered a more conditional take. “It’s not a scientific or engineering question,” he said. “If the cost of electricity on Earth drops a little bit. then it’s vastly superior to build [data centers] on Earth. If the cost of electricity remains the same, then it is possibly more economical to build them in orbit.”.

At the center of the IPO story is another truth: SpaceX’s biggest business today is not Starship or orbital data centers. It’s Starlink.

In 2025, SpaceX made nearly $4.1 billion from its space segment, at an operational loss of $657 million. But it made almost $11.4 billion from its satellite broadband service, Starlink, which remains by far the company’s biggest revenue generator.

SpaceX has more than 9,600 Starlink satellites in orbit, accounting for roughly two-thirds of all active satellites circling Earth. The company is planning for a total of 42,000. About three-quarters of SpaceX missions last year were for Starlink alone.

Starlink now has 10.3 million broadband subscribers in 164 countries, up from 8 million in November, according to the company’s prospectus.

With Starlink, SpaceX disrupted terrestrial internet providers and the niche industry of satellite television. In its next phase of growth, it is tackling mobile connectivity, which Bruegger calls “a bigger market by several orders of magnitude.”

SpaceX’s vertical integration is one reason the company is drawing attention. Through partnerships with more than 30 mobile network operators. SpaceX’s direct-to-device cellular service. Starlink Mobile. provides data. voice. and messaging to approximately 7.4 million monthly unique devices in about 30 countries. eliminating coverage “dead zones.” Since January 2025. Starlink has grown its V1 Mobile satellites from 360 to approximately 650. Starship will be able to deploy approximately 50 mobile satellites per launch, accelerating expansion.

SpaceX’s prospectus claims a total addressable market of $1.6 trillion for Starlink, including $740 billion for Starlink Mobile.

But competition is intensifying in the satellite-to-smartphone race. AST SpaceMobile now delivers direct-to-device cellular service using larger satellites that could let it scale quickly. In April 2026. Amazon Leo acquired Globalstar for $11.5 billion. giving it licensed L-band and S-band spectrum—critical frequencies for connecting to phones on the ground.

Defense is another steady pillar in the space economy. SpaceX’s first two Falcon 1 launches were purchased by the U.S. Department of Defense. The company now holds a commanding position as space-defense budgets rise.

The Trump administration has requested $71 billion in total Space Force funding for FY 2027. a 77% increase over 2026. including $40 billion designated for R&D. SpaceX is the leading launch contractor for the Space Force’s program focused on high-priority military and intelligence payloads; last April. it secured a $5.9 billion award for 28 launches through 2029.

ULA was awarded nearly $5.4 billion for 19 launches of its Vulcan Centaur, while Blue Origin received $2.4 billion for 7 missions. SpaceX also stands to earn another $2 billion in satellite work tied to the Golden Dome defense system.

Rocket Lab has its own defense contracts: U.S. Space Development Agency contracts worth a combined $1.3 billion to design and build secure communications and missile-warning satellites. plus a $190 million hypersonic-test launch contract. Earth monitoring satellite companies such as IceEye and Planet Labs have also seen revenues surge on defense demand. though their civilian business lines are marginal in comparison.

In the end, the IPO is being sold as a turning point for the industry. SpaceX’s first major exit is expected to catalyze the sector. releasing a wave of capital and experienced talent as early investors take their payouts and longtime employees cash in their equity and fan out through the broader space ecosystem.

Beck of Rocket Lab sees the IPO as “net positive” for the industry—and a potential boon to his company’s market cap. “As kind of the number-two behind SpaceX, we will be the comp. I’d take some of their multiples,” he said.

He also offered a blunt warning about what a bad outcome could signal. In the unlikely event that the SpaceX IPO is a flop, Beck said, “space will become less interesting for everybody, as well. But Elon doesn’t generally have a habit of shitting the bed, so I think everything will be fine.”

Public markets, he argues, will change the temperature inside the space industry. Beck described the sector as a place where “a smart CEO can wave their arms and tell a story. and unless you’ve got a PhD in mathematics. you can’t prove or disprove it. That’s led to some large failures.”.

Public-market analysts asking direct questions will make it harder to sell purely speculative ambitions. “That’s good for the industry, even if it’s uncomfortable for some of the players in it,” Beck said. As head of a public company since 2021. he says he is used to answering to analysts: “It’s basically super-easy. my job: Tell the truth. do what you say you’re going to do. and hit your numbers.”.

The tension sits right where the IPO pitch becomes most fragile. Running a public space company. Beck said. also means making sure that if delivery slips beyond promised timeframes. the rest of the business doesn’t suffer. SpaceX appears to be doing the opposite—putting a massive chunk of its proposed valuation on how quickly Starship can become a reliable ride for the company’s business arms.

As the company’s prospectus notes, any failure or delay in the rocket would hamper SpaceX’s “deployment of next-generation satellites, global satellite-to-mobile connectivity, and orbital AI compute.”

SpaceX’s ambitions are sky-high. But the economics of orbit still come down to a simple question that investors can’t waive away: when the roadmap depends on thousands of launches and a new rocket family performing on schedule, what happens if the world’s bottleneck doesn’t loosen fast enough?

SpaceX IPO Starship Starlink xAI launch capacity bottleneck orbital AI data centers FCC deadline Rocket Lab Neutron ULA Vulcan Centaur Blue Origin New Glenn satellite broadband defense contracts market valuation