Tech • AI • Robotics • Game
🤯🤯🤯 Elon on 39x SpaceX Revenue Surge by 2028
9.4/10TL;DR
SpaceX investor Shaun Maguire has published a model projecting $731 billion in revenue by 2028, driven mainly by a massive AI compute business, and Elon Musk replied: “I think we can do better.”
KEY POINTS
A forecast far beyond current scale
Maguire, a Sequoia Capital partner and major SpaceX backer since 2020, modeled companywide revenue of about $51.5 billion in 2026, $273.8 billion in 2027, and $731 billion in 2028. That would imply a jump of roughly 39 times compared with SpaceX’s reported 2025 revenue base. The forecast is framed as a 50th percentile case rather than an upside scenario.
Most of the growth comes from compute
The model’s central claim is that SpaceX is becoming an infrastructure provider for large-scale AI or superintelligence compute, not just a launch and satellite company. Maguire assigns about $569 billion of 2028 revenue to terrestrial compute, roughly 78% of the total. He argues demand is being pulled by companies willing to pay extraordinary sums for capacity because additional compute can translate quickly into higher product usage and revenue.
Claims of major customer demand
The thesis assumes that top AI developers will keep leasing vast amounts of capacity as the race to train and run advanced models intensifies. Maguire cited expectations that Anthropic could be paying more than $3 billion per month by year-end for access to capacity, with Google near $1 billion per month and another customer at a similar level. Those figures underpin the argument that compute pricing can remain high even as supply expands.
Why supporters say the model is not arbitrary
Backers of the forecast point to Maguire’s earlier Starlink projections from 2020, when the service had only begun commercial operations. That model estimated 2025 Starlink revenue at $18.5 billion and 2026 at $25.2 billion. Actual 2025 revenue was cited at $11.38 billion, with 2026 tracking closer to the earlier estimate, suggesting the forecast was aggressive but directionally strong.
Starship is treated as the core moat
A major pillar of the case is Starship, which Maguire describes as a historic competitive advantage because it could slash launch costs and accelerate deployment of satellites, communications hardware and orbital infrastructure. He argues most of the world is still trying to match capabilities SpaceX demonstrated with Falcon 9 years ago, while Starship represents another step change in payload and economics.
Low-cost producer logic
Maguire compares AI compute to a commodity market where the lowest-cost operator can outlast rivals through price pressure. His argument is that if compute prices fall, weaker providers suffer first, while SpaceX can consolidate share because its capital costs, logistics and engineering execution are stronger. He says revenue per gigawatt and capital expenditure per gigawatt should decline together, cushioning margin pressure.
Starlink and direct-to-cell still matter
The long-term bull case is not only about compute. Maguire also argues that Starlink and direct-to-cell services remain underappreciated, especially once Starship is flying at scale. He notes the current Starlink network was effectively built with around 400 Falcon 9 launches, while one fully loaded Starship mission could equal roughly 20 Falcon 9 launches in payload terms.
Economics per launch could be enormous
Based on that launch capacity, Maguire argues a single Starship mission could support around $1 billion in annual revenue and about $5 billion in lifetime revenue over five years across Starlink, direct-to-cell or orbital compute use cases. He further suggests direct-to-cell could surpass core Starlink broadband revenue by 2029 or 2030 because the addressable market is much larger.
Skepticism remains intense
Critics have mocked the model as unrealistic, especially the implied profitability and the idea that a company known for rockets could rapidly build a compute business on the scale of several cloud giants. Some comparisons focused on whether the forecast effectively assumes SpaceX can create multiple Amazon Web Services-sized businesses in a few years with unusually high margins. Maguire’s response is that traditional cloud firms are far weaker at building physical infrastructure quickly.
Musk’s brief endorsement amplified the debate
The most consequential response came from Musk, whose six-word reply signaled that even the already extreme forecast may be conservative in his view. That reaction has intensified attention on whether SpaceX should now be valued primarily as a launch company, a communications network, or a future compute and infrastructure giant.
CONCLUSION
The debate over SpaceX now centers less on rockets alone and more on whether it can dominate the physical buildout of the AI economy. If Maguire’s thesis is even partly correct, the company’s next growth phase could be far larger than the market currently assumes.
Ask a question