SpaceX has more annualized cloud revenue than Google and is will run ahead of Amazon AI for AI cloud revenue. SpaceXAI has 20% of all active and completed AI data centers in the past two years. The lead in AI Cloud revenue is a $26 billion runrate from the Anthropic and Google deals.
This is existing revenue and existing ai data centers. SpaceX will add 220,000 B300 GPUs that will replace all of the data center they are renting to Anthropic and Google.
SpaceX still has 60% of its colossus 2 data center to operate Cursor and train all Grok AI models.
Renting out all of Colossus 2 would generate $55 billion per year and it cost them $40 billion to build it and they built it in 12 months.
In the next 12 months SpaceX will have five times Coreweave’s 2025 revenue and will have double the projected revenue for 2026. SpaceX can still double its revenue if they rented all that they have currently already built.

Tesla can also activate this revenue stream for existing and near term AI data centers and distributed AI.

So Google and Anthropic combined will alone bring $26bn in annual revenue for xAI
Remember, before these 2 deals xAI was valued at $250bn (at the time of the SpaceX merger). It's revenue was $3.2bn. It's on track for $30bn this year
But note the breakup of the current $30bn… pic.twitter.com/P6Wkk2tB5W
— Flexible Capital ⭐ (@flexifinHQ) June 6, 2026

Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
Hélice Brian , did they plan other data center ? Seems a very smart investment
Major brokers have large difference in the value of this IPO. I have seen one as low as $785 billion, down by 55%.
Others have pointed out that only 5 percent of the stock will be released and this low float, plus I think I heard S&P changed their rules on joining it just for them, will cause a feeding frenzy in a couple of weeks with it popping and then going down to under $1 trillion.
Hm… It’s a very clever pivot from Elon given som negative conditions. Those negative conditions is that it’s more profitable to rent out their AI compute, than to use it to train their own models and/or run inference for their own models.
A cynic would say that SpaceX has all this spare capacity because no one is using X or grok.
That said, it does look like the suspicious AI pumping has arrived just in time for the SpaceX IPO. Google is renting SpaceX server space now, at some cost, without obviously needing to. The contract is also very generous in its cancellation terms, with a 90 days termination notice for both sides. Given the billions at stake, thats highly unusual. Furthermore, Alphabet owns around 6% of SpaceX.
Brian writes some vety interesting articles, but those numbers are off the chart, too big objectively. I think AI is hitting a plateau. They released opus 4.8 and still no ratings are published on lm arena. Next gemini 3.5 still not released, grok 4.3 was not so good. They released gemini 3.5 flash but compared with 3.0 flash the improvement is tiny. Tesla builds things fast and I get the logic of getting rich by services to “gold miners” in AI rush time, but I still see too big numbers.
Hélice Brian , did they plan other data center ? Seems a very smart investment