![\"Illustration:](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/06\/d1fb5ae8aa9c7e487f86b70db5fd595b.gif\")
Illustration: Brendan Lynch\/Axios <\/p>\n Illustration: Brendan Lynch\/Axios <\/p>\n
It’s a foregone conclusion that SpaceX will raise at least $85 billion in its record-wrecking IPO<\/a>, with pricing set for Thursday and first trades for Friday.<\/p>\n The bull case <\/p>\n SpaceX will generate hundreds of billions of dollars in revenue by 2030, despite booking less than $19 billion last year<\/a>.<\/p>\n The biggest chunk should be from Starlink, which will leverage Starship’s larger payloads to supercharge its nascent direct-to-cell business. Starlink could corner the global market, either as a replacement for legacy carriers or as their new infrastructure layer (thus reducing consumer switching costs). Don’t forget about its massive spectrum purchase<\/a> from Echostar.<\/p>\n The floor for SpaceX’s AI business is that it just keeps selling compute, like the new agreements with Anthropic and Google \u2014 which combined will provide around $2 billion in monthly revenue.<\/p>\n The ceiling for SpaceX’s AI business is that Grok really works as a viable alternative to Claude, ChatGPT, and Gemini (plus their various offshoots). Then the sky’s the limit, although SpaceX may need to reclaim some of its compute.<\/p>\n SpaceX also will continue its flagship launch business, including some lumpiness tied to NASA’s lunar missions, and possibly could be generating revenue from orbital data centers where revenue horizons are longer.<\/p>\n If SpaceX hits around $200 billion in revenue with continued growth, a $1.75 trillion valuation will look like a bargain.<\/p>\n The bear case <\/p>\n SpaceX has spent most of its life operating as the market leader in a category of its own making (commercial space launch).<\/p>\n Starship remains a work in progress, not a sure bet. Coverage may be superior to current plans, but will it be cost competitive? Starlink’s current service is adding subscribers, but at a declining ARPU. There also may be viable rivals \u2014 including Amazon, which is buying Globalstar<\/a>.<\/p>\n Compute is a commodity. Right now it’s benefiting from the bottleneck, but that should ease given the massive number of data center buildouts and expected increases in inference efficiency. And if Grok does work, it’s possible that rivals like Anthropic and Google would find ways to cut off SpaceX’s compute cash cow.<\/p>\n The wildcard: Elon Musk <\/p>\n