SpaceX reported its first quarter as a public company after the close on August 4. Revenue was $7,814M, up 92% year over year and 66% sequentially, against a Street expectation near $6.72B. The net loss was $541M, or $0.09 a share, against an expected loss of $0.26. Adjusted EBITDA was $3,538M, nearly triple the year-ago $1,214M.
The stock fell as much as 8% in after-hours trading, from a $125.33 close toward $116.70. Our preview on July 31 named the AI segment's losses as the thing the print had to address. It addressed them. The market sold it anyway, and the reason is on the capex line.
What we said to watch, answered
| What we said to watch | What the print says |
|---|---|
| Whether the AI segment's losses are narrowing or widening | Narrowing sharply. Operating loss $1,257M against $2,469M in Q1 — cut 49% — on revenue up 213% sequentially. |
| Starlink economics holding as ARPU falls | Held. Connectivity revenue $4,291M, up 66% YoY and 32% sequentially, with segment operating income up 79%. |
| Whether the market is right to value the AI segment near zero | Harder to argue now. AI revenue $2,561M, up 247%, with positive segment adjusted EBITDA of $1,146M against −$276M a year ago. |
| Starship as the physical proof point | Two successful V3 flight tests in the past 90 days; Space segment revenue $962M, up 55% sequentially. |
The segments
| $M | Q2 2026 | Q1 2026 | Q2 2025 |
|---|---|---|---|
| Space revenue | 962 | 619 | 746 |
| Connectivity revenue | 4,291 | 3,257 | 2,588 |
| AI revenue | 2,561 | 818 | 737 |
| Total revenue | 7,814 | 4,694 | 4,071 |
| Space operating income | (542) | (662) | (369) |
| Connectivity operating income | 1,656 | 1,188 | 923 |
| AI operating income | (1,257) | (2,469) | (1,524) |
| Total operating loss | (143) | (1,943) | (970) |
| Capex | 18,369 | 10,107 | 2,825 |
Three facts sit in that table and they do not point the same way.
The first is that Connectivity is the company. It produced $1,656M of operating income while the consolidated result was a $143M operating loss — Starlink is funding the other two segments outright, and the preview's framing of a profitable connectivity business carrying an unprofitable AI business survived contact with the numbers.
The second is that the AI segment stopped being a pure cost. $2,194M of the $2,561M is AI solutions and infrastructure revenue, up from $475M in Q1, driven by new Cloud Services Agreements — SpaceX says $1.6 billion of the quarter's AI infrastructure revenue was incremental from those deals, against $14.1 billion of contracted sales signed. Advertising revenue, the X platform's contribution, went the other way: $367M against $426M a year ago.
The third is the one that moved the stock. Capital expenditure was $18,369M in a single quarter, of which $15,828M was the AI segment — up from $749M a year ago. SpaceX spent more on AI infrastructure in one quarter than it earned in revenue across all three segments.
Why this site cannot give SpaceX a Rule of 40 score
The Rule of 40 needs two inputs: revenue growth and free-cash-flow margin. We have the first — 92%, the highest of any company on this site this season. We cannot compute the second, and the reason is worth stating rather than papering over.
Our free-cash-flow definition is operating cash flow minus capital expenditure, quarterly. SpaceX's release publishes selected cash-flow information for the six months only: operating activities of $3,466M year to date. It publishes capex per quarter and per segment, so the capex half is known. The operating half is not, because Q1 2026 predates the June 12 IPO — there is no earlier quarterly report to subtract, and none will ever exist.
So the honest arithmetic is the half-year one: $3,466M of operating cash flow against $28,476M of capex is −$25,010M of free cash flow in the first six months of 2026, a −200% margin on $12,508M of half-year revenue. That figure is real, it is enormous, and it is not a quarterly number, so it does not go into the series or produce a score.
What we have stored is what the release supports: revenue for both quarters, gross margin for both, and diluted EPS for the quarter SpaceX reported as a public company. The free-cash-flow series stays empty until a second quarterly report makes the operating-cash-flow line differenceable — and the 0 that SpaceX's stock page still shows in its Rule of 40 badge is that absence rendering as a number, not a score the company earned. The alternative — annualising, or splitting the half-year evenly across two quarters — would produce a number that looks like the other 65 companies' numbers and is not one.
Gross margin, which we can compute
Cost of revenue was $3,495M on $7,814M of revenue, a 55.27% gross margin, against 49.13% in Q1. Six points of gross margin in one quarter, on 66% sequential revenue growth, is the operating leverage the CFO commentary claims, and it is checkable from the income statement rather than taken on the word of a non-GAAP measure.
Note what sits below that line: research and development of $3,548M — 45% of revenue, more than the entire Space segment's revenue, and larger than the reported operating loss by a factor of twenty-five. This is a company whose GAAP loss is small because gross profit is now large enough to nearly cover an R&D budget that would be the whole story at most companies.
The balance sheet is the thesis
SpaceX ended the quarter with $93.5B of cash and equivalents plus $6.5B of marketable securities — the $100 billion the CFO cited — against $39.4B of debt and finance leases, and $47.5B of backlog. Total assets went from $92.1B to $192.8B in six months, with property, plant and equipment alone rising from $42.6B to $65.7B.
The IPO raised most of that: financing activities brought in $100.3B in the half. Which reframes the capex number rather than excusing it — SpaceX is spending IPO proceeds on AI infrastructure at roughly $18B a quarter, and at that rate the $100B lasts about five and a half quarters before the business has to fund itself or return to the market. The announced $60 billion acquisition of Cursor is against that same balance.
The bottom line
The preview asked whether the AI segment's losses were narrowing or widening, and said the market was pricing that segment at close to nothing. Both parts got an answer: the loss halved while revenue tripled sequentially, and pricing that at zero is now a harder position to hold.
The after-hours reaction says the market swapped one worry for another. On the numbers, that is defensible — a company that halves an operating loss while quadrupling its capital expenditure has not become cheaper to own, it has changed what you are underwriting. What it has not done is give anyone enough cash-flow disclosure to judge that on a quarterly basis. The next report is the one that makes the second half of our score computable, and we will publish it the day it does.
SpaceX figures are from the company's Q2 2026 results release, published August 4, 2026 and captured in our release notes: revenue $7,814M, cost of revenue $3,495M, R&D $3,548M, loss from operations $(143)M, net loss $(541)M, basic and diluted loss per share $(0.09) on 5,864 million weighted-average shares, adjusted EBITDA $3,538M, six-month operating cash flow $3,466M, quarterly capex $18,369M and six-month capex $28,476M, cash and equivalents $93,522M plus marketable securities $6,487M. Q1 2026 comparatives are the ones disclosed in the same release, not a separately filed quarterly report. Gross margin is revenue less cost of revenue, the same derivation used across this site. After-hours prices are as reported on the day of the print and are not stored in this repo. Rule of 40 definition: our explainer.