SpaceX Just Became the Biggest IPO in History — and xAI Is Riding Inside It. If You Build on Grok, Your Dependency Just Got 1,000 New Shareholders.
SpaceX starts trading on Nasdaq today under the ticker SPCX, priced at $135 a share, valuing Elon Musk's rocket-and-Starlink-and-xAI group at roughly $1.75 trillion and raising about $75 billion. That's the largest IPO ever recorded, by a wide margin. Most of the coverage you'll see today is about the rocket company and the index funds that now have to buy it. The part that matters if you're a builder: xAI is inside this thing, and the company that makes Grok just swapped venture money for public-market money.
I wrote about this back when the S-1 dropped, and the headline number then was that xAI was burning around $2.5 billion a quarter. That number didn't go away. It just got a much bigger balance sheet behind it. If you have Grok anywhere in your stack (as a primary model, a fallback route, or a cheap-tokens experiment), today is the day to think clearly about what changed.
What actually happened
The mechanics are worth getting right because the hype version is sloppy.
SpaceX priced its IPO at a fixed $135 per share and debuts on Nasdaq today under SPCX. The deal values the combined entity at about $1.75 trillion and targets roughly $75 billion in proceeds. xAI isn't a separate ticker. It merged into SpaceX in February 2026 in a $250 billion all-stock deal, which is how Grok ended up riding inside a launch-services company in the first place. After that merger, SpaceX was valued around $1.25 trillion; today's listing marks it up further.
There's one more piece that creates real near-term pressure. Under a Nasdaq rule change from March 2026, a mega-IPO this size can be fast-tracked into the Nasdaq-100 within 15 trading days instead of waiting the usual months. That means a wave of forced buying from every index fund that tracks the Nasdaq-100, on a compressed clock. It's good for the stock and irrelevant to your code, but it tells you how much institutional money is about to be mechanically tied to this company's quarterly performance.
Why a funding change is a dependency change
Here's the thing solo builders underrate: the cap table of a model vendor is a feature of the product you're building on.
When xAI was a private company torching $2.5 billion a quarter, it could price Grok aggressively to buy market share, because that's what venture-funded land-grabs do. Losses were the strategy. Nobody had to answer for them every 90 days. The moment that company sits inside a public entity, the incentive structure shifts. Public companies get graded quarterly, and "we're losing billions to undercut Anthropic and OpenAI on token pricing" is a harder story to tell a shareholder base that includes index funds and pension money than it is to tell a late-stage venture round.
That cuts two ways, and honestly both directions are real.
The optimistic read: a $75 billion raise is the most stable Grok has ever been. A war chest that size means the roadmap doesn't depend on the next fundraise, the compute keeps flowing, and the model you're calling isn't going to get orphaned because a private round fell through. For a dependency, "bankrolled by the largest IPO in history" is not the worst thing to be.
The cautious read: aggressive subsidized pricing has a shorter shelf life inside a public company. If part of why you're on Grok is that it's cheap, understand that the cheapness was a venture-stage decision, and venture-stage decisions don't survive contact with an earnings call forever. The price you're paying today is not a promise.
What this means for your stack
I'm not going to tell you to rip Grok out. That would be exactly the kind of reflexive vendor-migration advice I think is mostly wrong. What I'd do instead is treat today as a forcing function to check three things.
First, know how load-bearing Grok actually is for you. If it's a fallback route or one option in a model router, this is a non-event: you already have optionality, which is the whole point of routing. If it's the single model your product can't function without, you have concentration risk, and that risk just got a news hook.
Second, separate the price question from the capability question. If you're on Grok because it's genuinely the best model for your specific task, that's a real reason and a public listing doesn't change it. If you're on it mostly because the tokens are cheap, write down what your unit economics look like at, say, 1.5x or 2x the current price, so a future pricing change is a spreadsheet update and not a panic.
Third, make sure switching is cheap before you ever need to. The protection against any single model vendor's corporate drama isn't loyalty and it isn't paranoia. It's an abstraction layer thin enough that swapping a model is a config change. If you've got that, the SpaceX IPO is interesting trivia. If you don't, today is the nudge to build it on a calm afternoon instead of during an incident.
The honest counter-take
It's very possible none of this touches you for a long time. Public companies can and do keep subsidizing strategic products for years: Amazon ran AWS-adjacent bets at a loss for ages, and Musk-run companies in particular have a long history of telling public shareholders to be patient and getting away with it. xAI's pricing might not move at all in 2026. The merger might turn out to be a pure capital-access play that leaves the model roadmap untouched.
So this isn't "Grok is about to get expensive, flee." It's narrower than that: a vendor in your potential stack changed how it's funded and who it answers to, and that's exactly the kind of quiet structural shift that's easy to miss because it shows up as a finance headline instead of a changelog.
What I'd actually do
If Grok is in your stack, spend twenty minutes today doing one thing: confirm you could swap it for another model with a config change, and write down your costs at 2x current pricing. That's it. If both of those are already true, you're done and you can enjoy the rocket headlines. If either one isn't, you just found the actual work, and it's an afternoon now versus a fire later. The largest IPO in history is a great reminder that the models you build on are companies, and companies change shape. Build like that's true, because it is.
Author
Lukas
@lukcombinator