# SpaceX's record IPO just put a frontier AI lab on the public market

> The $1.77 trillion debut is the largest IPO in history. The detail that matters more sits in the prospectus: SpaceX now owns xAI, so the float quietly took Grok public, and that changes what you're betting on.

- **Pillar:** News
- **Author:** Aditya Marin Gasga (Founding Editor)
- **Published:** 2026-06-12T20:27:00.000Z
- **Tags:** spacex, xai, grok, ipo, ai business

## TL;DR

SpaceX priced its IPO at $135 a share for a $1.77 trillion valuation and about $75 billion raised, the largest IPO in history, and closed its first day up roughly 19% at $161.11. Because SpaceX absorbed xAI in February 2026, that float also took Grok and X public, so a frontier AI lab now sits inside a company answering to quarterly earnings, which matters more for anyone building on xAI than the valuation does.

## Key takeaways

1. SpaceX priced at $135 a share for roughly $75 billion raised and a $1.77 trillion valuation, the largest IPO ever, ahead of Saudi Aramco's $25.6 billion in 2019.
2. SPCX opened at $150 on Nasdaq, touched $176.52, and closed its first session at $161.11, up about 19%, for a market cap near $2.1 trillion.
3. The prospectus folds in xAI, acquired effective February 2, 2026: SpaceX's 'AI segment' now includes its AI compute, Grok, and X.
4. Grok is now inside a public company, and its content, accuracy, and 'misinformation and deepfakes' exposure are named in SpaceX's own risk factors.
5. For teams building on xAI, the model didn't change; what changed is that the vendor now optimizes for quarterly earnings and carries public-company disclosure risk.

import SignalChart from '~/components/viz/SignalChart.astro';
import PullQuote from '~/components/article/PullQuote.astro';

SpaceX priced its IPO at $135 a share on Thursday and began trading Friday on Nasdaq as SPCX, raising about $75 billion at a [$1.77 trillion valuation](https://www.cnbc.com/2026/06/03/spacex-ipo-stock-price-roadshow-musk.html). That makes it the largest IPO in history, bigger than Saudi Aramco and Alibaba combined. The number everyone will quote is the valuation. The number that matters more is in the prospectus: SpaceX now owns xAI, so the float just took Grok public.

Strip out the rocket-company story and what is left is the first time a frontier AI lab has been folded into a public company's balance sheet. Grok and X are line items now, and line items answer to quarterly earnings. For anyone building on xAI's models, that shift changes the calculation more than the headline number does.

## What actually happened

SpaceX sold [555,555,555 Class A shares](https://www.sec.gov/Archives/edgar/data/1181412/000162828026040364/0001628280-26-040364-index.htm) at a fixed $135, raising roughly $75 billion at a $1.77 trillion valuation ([Fortune](https://fortune.com/2026/06/03/spacex-ipo-share-price-index-funds-valuation-public/)). The stock opened at $150, ran as high as $176.52, and [closed its first session at $161.11](https://www.cnbc.com/2026/06/12/spacex-ipo-spcx-live-updates.html), up about 19% from the offering price, for a market capitalization near $2.1 trillion. Elon Musk, who keeps voting control through a dual-class structure, [became the world's first trillionaire](https://www.cnn.com/2026/06/12/business/live-news/spacex-goes-public-ipo) on paper by the bell.

The "largest ever" line is not marketing. Saudi Aramco raised $25.6 billion in 2019. Alibaba raised $21.8 billion in 2014. SpaceX raised roughly three times either of them in one offering.

<SignalChart
  type="bar"
  orientation="horizontal"
  pillar="news"
  categories={["SpaceX (2026)", "Saudi Aramco (2019)", "Alibaba (2014)", "SoftBank (2018)", "NTT DoCoMo (1998)", "Visa (2008)"]}
  categoriesShort={["SpaceX", "Aramco", "Alibaba", "SoftBank", "DoCoMo", "Visa"]}
  series={[
    { name: "Amount raised", data: [75, 25.6, 21.8, 21.3, 18.4, 17.9], accent: "news" },
  ]}
  prefix="$"
  suffix="B"
  title="Largest IPOs ever, by amount raised"
  caption="SpaceX's ~$75B raise is roughly triple the prior record. Figures are primary proceeds, rounded."
  source={{ label: "Visual Capitalist", href: "https://www.visualcapitalist.com/ranked-the-biggest-ipos-in-history-spacex/" }}
/>

A debut this large pulls index funds in behind it, which is a story of its own. We walked through the mechanics of a giant AI-adjacent listing when [OpenAI filed its own trillion-dollar IPO](/openai-trillion-dollar-ipo-explained). This one carries a wrinkle that listing did not.

## Why a rocket company owns a frontier AI lab

SpaceX acquired xAI in an all-stock deal that took effect [February 2, 2026](https://www.cnbc.com/2026/02/11/musk-announces-xai-re-org-following-key-departures-spacex-merger.html), and the prospectus recasts the financials to fold xAI in. The filing names an "AI segment" that, in its own words, "includes our AI compute, Grok, and X," and describes the combined business as a vertically integrated company spanning launch, satellite internet, AI, and social media.

That structure is the whole point. The February deal reportedly valued xAI at roughly $250 billion as a standalone AI business ([CNBC](https://www.cnbc.com/2026/02/11/musk-announces-xai-re-org-following-key-departures-spacex-merger.html)), a level that makes raising frontier-scale compute hard to sustain alone. Inside SpaceX, Grok rides a balance sheet with $18.7 billion in 2025 revenue and the cash from the largest IPO on record. The compute bill that pins down every independent lab just got easier for this one to pay.

It also means you can no longer hold SpaceX's rockets without holding a chatbot's moderation problem. The risk factors say as much: SpaceX's "social media and AI-related activities expose us to a variety of risks related to harmful, misleading or illegal content, accuracy, misinformation and deepfakes, bias, discrimination, toxicity." That sentence now sits in a filing read by every institutional investor who buys the stock.

## What changes if you build on Grok

Here is the counter-take, and it runs against the easy read. The easy read is that a better-capitalized xAI is a safer vendor. That is half right. The capital is real and the runway is longer, so the risk that xAI simply runs out of money to serve your API calls dropped on Friday.

The other half is what public ownership does to incentives. A private lab can lose money on a model for years to win developers. A segment inside a quarterly-reporting company answers to that segment's contribution to the next earnings call. Pricing that looked like a land grab can tighten. A model that does not pull its weight can lose internal funding to the parts of SpaceX that do. And every Grok controversy becomes a disclosure event rather than a private embarrassment, which nudges a frontier lab toward the caution that quietly reshapes what a model will and will not say.

<PullQuote pillar="news">What changed isn't Grok's quality. It's who Grok answers to, and the answer is now public shareholders on a ninety-day clock.</PullQuote>

None of this touches Grok's capability today. The weights are the weights, and the API behaves the way it did Thursday. If you chose xAI for what the model can do, you keep it for what the model can do.

## What to do about it

If xAI sits in your stack, nothing breaks Monday, but two things deserve a calendar note. Watch the first earnings call for how SpaceX reports the AI segment, because segment-level disclosure will tell you fast whether Grok is being run for growth or for margin. And keep a fallback model qualified, the ordinary hygiene for any single-vendor dependency, because "too big to disappear" and "stable to build on" are not the same property. The IPO settled the first one. It did nothing for the second.