In just a couple of weeks, SpaceX will IPO.

For investors, this is a watershed event. First and foremost, its estimated market capitalization of $1.75 trillion will make SpaceX the biggest IPO in history. Such an event “feels” like it’s inaugurating a new era in investing — the era of investing in space — and we’re seeing space stocks of all sorts leaping higher as the excitement mounts.

At the same time, the SpaceX IPO will test valuations in the space sector, give investors clear insight into how the biggest company in space is performing, and potentially yield insight into how the rest of the sector might evolve as smaller space stocks “grow up” to SpaceX-size.

Helping give us this insight is the SpaceX IPO prospectus, which SpaceX unveiled publicly last week. Here are three things you need to know about it.

1. Official numbers are in

Just weeks ago, investors contemplating the SpaceX IPO still had to rely on guesses about SpaceX’s financial health to determine how much SpaceX stock might be worth.

As it turned out, research outfits such as Payload Space did a great job guesstimating SpaceX’s 2025 revenue at $15 billion to $15.5 billion, while Reuters was close to the mark, reporting that SpaceX was profitable in 2025 (at least, before absorbing money-losing X and xAI).

But with the prospectus in hand, we can now see the official 2025 numbers for SpaceX’s three main divisions and the company as a whole. Here they are:

Data source: SpaceX IPO prospectus; all numbers in millions. SpaceX earned a net profit of $791 in 2024. SpaceX’s net loss in 2025 was $4,937.

Space

SpaceX grew Falcon launches from 134 in 2024 to 165 in 2025. Most of these were revenue-less Starlink launches, though, so revenue grew only 8% despite 30% more launches.

Worse, SpaceX’s Space business flipped to a 2025 loss as the company poured research and development dollars into developing the new Starship rocket.

Connectivity

Starlink’s customer count doubled year over year. Revenue increased only 50% because most of this growth occurred outside the U.S., where SpaceX charges lower prices. Still, revenue grew twice as fast as costs, allowing profits to more than double.

AI

Revenue rose 22% year over year as X and Grok subscriptions grew. Cost of revenue grew faster at 29%; however, expensive AI chips and cloud computing services ate into gross margin. R&D costs exploded — up 331%. Even selling, general, and administrative costs outgrew revenue — up 65%.

Long story short, we all knew artificial intelligence (AI) was a money pit. Now we know how deep it is.

2. Starlink is thriving; everything else is not so much

Now, let’s zoom in on SpaceX’s most recent results. Comparing Q1 2026 revenue to Q1 2025, we see that SpaceX’s Connectivity (Starlink) business is still growing nicely. Q1 revenue of $3.3 billion grew 32% year over year.

Space revenue, however, declined 28% year over year. AI revenue grew 12% — which in any other industry would be quite respectable, but in the white-hot field of artificial intelligence seems rather tame. These increasing losses at Space and AI may make IPO investors nervous.

Data source: SpaceX IPO prospectus, all numbers in millions.

Space

So what’s going on here? In Space, SpaceX blamed a decrease in customer launches for falling revenue and “accelerated investment” in Starship for growing losses. Broadly speaking, this was the same dynamic seen in 2025 — but more so.

Connectivity

The story with Starlink is likewise similar to what we saw in 2025. Here, SpaceX credited 105% growth in Starlink subscribers — even faster than 2025 growth — for its 32% increase in revenue. Yet customer growth and revenue growth are still diverging as non-U.S. growth, where Starlink prices are lower, predominates.

More concerning is the steep slowdown in profit growth. As SpaceX spends more on replacing “dead” Starlinks and less on growing its constellation, I’d expect this to put a ceiling on profitability. But the bigger Starlink gets, the more pricing power it may have — allowing SpaceX to raise prices and improve profits down the road.

AI

SpaceX credited increased Grok and X subscriptions for revenue growth in this segment, plus a bit of money from “data licensing agreements,” offset by weaker advertising revenue on X.

The bad news: AI is losing money faster than ever.

3. Things we don’t know

And now for the other bad news: Even at this late date, SpaceX’s prospectus still contains many blank spaces. We don’t know how many shares SpaceX will sell. We don’t know how many shares of SpaceX exist.

We don’t know the price at which SpaceX will IPO, either, or what the company’s market capitalization will be. Media reports guess it will be somewhere between $1.5 trillion and $2 trillion, but until all the blanks are filled in, we don’t really know.

Finally, we don’t know how many SpaceX shares will be “Class A” and how many will be “Class B.” We do know that “each share of Class A common stock will entitle its holder to one vote per share,” while “each share of Class B common stock will entitle its holder to 10 votes per share,” granting Elon Musk a controlling interest in SpaceX for eternity. As the prospectus admits:

Upon completion of this offering, Mr. Musk will beneficially own a majority of the voting power of our common stock and the Class B common stock, which elects a majority of the board. As a result, we expect to be a ‘controlled company’.

So ultimately, whether SpaceX succeeds or fails will depend on Elon Musk.

Just as it always has.