
Spending Like a Hyperscaler

SpaceXAI reported $18.37 billion in capex for its first public quarter, with $15.83 billion dedicated to AI infrastructure — beating the $13.2 billion consensus. In the same quarter, Amazon spent $53.
1B, Alphabet $44.9B, Microsoft $41.0B, and Meta $31.1B. The sequential dollar increases are comparable across all five firms.
The critical difference lies in funding: operating cash flow covers 155% of capex at Microsoft and 106% at Meta, but only 12% at SpaceXAI.
Oracle covers 89%, Alphabet 87%, Amazon 84%, and CoreWeave 39%.
Equity and credit markets have repriced SpaceXAI downward: its stock (SPCX) closed at $108, a fifth below the $135 IPO price and half its June peak.
Every tranche of June’s $25 billion debut bond trades below par — the 2031s at 98 cents on the dollar and the 2056s at 90 cents.
Despite the funding gap, SpaceXAI is pushing to compete: it has data centers, trained a model (GroK 4.
5) ranking fourth on the Intelligence Index within six points of the leader, and has millions of developers using Cursor.
The IPO gave SpaceXAI a cost-of-capital advantage, but the question is whether that advantage can last long enough for the AI business to reach parity with the hyperscalers.


