Author | Zhiyu Learn more financial information | BT Finance Data Pass The full text comprises 2,322 words, with an estimated reading time of 9 minutes.

On May 21, 2026, SpaceX unveiled its prospectus.

Valued at a staggering $1.75 trillion, it is poised to make history as the largest-ever Initial Public Offering (IPO).

However, that's not the most striking aspect.

The prospectus unveils three pivotal figures that, when analyzed collectively, illuminate why Musk opted for this moment to go public.

Starlink: $11.4 billion in revenue in 2025, marking a 50% year-over-year surge—a veritable cash cow.

AI segment (integrating parts of xAI): $800 million in revenue in Q1 2026, accompanied by an operating loss of $2.469 billion—a significant drain on finances.

Starship rocket: Cumulative R&D investment surpasses $15 billion, yet substantial revenue generation remains elusive—a capital-intensive endeavor.

Three fronts, advancing in tandem. The cash inflow from Starlink cannot keep pace with the combined outflows from AI and Starship.

Musk had repeatedly vowed that SpaceX would remain private until a Mars colony was established.

Now, he has reneged on his promise. Because going public is no longer a mere option—it's a strategic imperative.

1. First, Starlink: SpaceX's Most Undervalued Asset

Starlink serves as SpaceX's financial linchpin.

According to SpaceX's prospectus, as of the end of March 2026, Starlink operates approximately 9,600 satellites in low Earth orbit, providing internet connectivity to around 10.3 million users.

In 2025, Starlink generated $11.4 billion in revenue, up 50% year-over-year. It is anticipated to sustain rapid growth in 2026.

Such growth rates are exceptional in any industry.

Yet, Starlink's strategic value transcends mere revenue. It holds a deeper significance: these 10.3 million users represent global pioneers in satellite internet—they have established subscription models, ground terminal ecosystems, and enterprise-level service systems, creating a formidable competitive barrier.

Any new entrant must not only launch satellites but also reconstruct this entire ecosystem.

China's Qianfan constellation is in hot pursuit, having just completed its ninth batch of network satellites on May 17. Nevertheless, it still lags significantly in terms of user scale and service system completeness.

2. Next, AI: Anthropic Pays a Hefty $45 Billion 'Entrance Fee'

SpaceX finalized its merger with xAI in 2024, officially integrating the AI segment into its operations.

The rationale behind this integration is clear: SpaceX operates the world's largest low Earth orbit satellite network, while AI demands computational infrastructure. Combining the two enables the construction of a "space-ground integrated AI computing" system—the cornerstone of the $28.5 trillion Total Addressable Market (TAM) outlined in the prospectus, of which $26.5 trillion pertains to AI.

However, the reality is that the AI segment generated only $818 million in revenue in Q1 2026, with a loss of $2.469 billion. Annual capital expenditures are projected to exceed $30 billion.

Then, just moments before the prospectus was made public, Yicai disclosed a crucial detail: Anthropic signed a cloud services agreement with SpaceX, committing to pay approximately $1.25 billion per month over the next three years, totaling around $45 billion.

What does this contract signify?

It indicates that Musk has devised a closed-loop system to offset AI's cash burn with AI computing revenue—building computing power and selling it to clients like Anthropic, using the proceeds to cover operating losses.

The $45 billion contract spans three years, equating to $15 billion in annual contract revenue, partially offsetting the AI segment's losses. However, with AI capital expenditures soaring, whether this offset is sufficient remains the critical question.

Analogy: It's akin to opening a restaurant that requires a massive initial investment and continues to operate at a loss each month. But you've just secured a three-year corporate catering contract—now the business truly takes off. SpaceX's AI segment is that restaurant on the ascent.

3. Finally, Starship: $15 Billion Invested, Anticipating the Next Breakthrough

Starship represents SpaceX's "mega-rocket" and the cornerstone of Musk's interplanetary ambitions.

The prospectus reveals that SpaceX has invested over $15 billion in Starship's development. It plans to commence launching next-generation Starlink satellites in the second half of 2026, carrying up to 60 per mission.

This step holds immense strategic value for SpaceX: the new Starlink satellites boast higher technical specifications and broader coverage. If Starship's launch costs stabilize, it will significantly reduce the marginal cost of Starlink network expansion, creating a "cheaper, denser satellite network"—the ultimate competitive advantage against rivals.

But until then, every launch represents a substantial R&D investment.

4. The Rationale Behind the Three-Pronged Strategy: Why Musk Must Go Public Now

Now, let's examine the three fronts collectively:

Starlink generates cash, but how much annually? Around $11.4 billion in 2025, projected to grow to $14–19 billion in 2026.

AI consumes cash, with capital expenditures expected to exceed $30 billion in 2026.

Starship requires capital, with cumulative investments surpassing $15 billion and continuing.

In aggregate, SpaceX's annual capital expenditure needs far outstrip the free cash flow generated by Starlink.

This is the true impetus behind Musk's decision to go public—he requires public market liquidity to sustain this "three-front war."

Moreover, he seeks not just a one-time financing but an opportunity to join the S&P 500 index. Once included, global funds tracking the index will passively purchase billions of dollars worth of SpaceX shares, providing sustained liquidity support.

5. A Noteworthy Detail in the Prospectus for Chinese Readers

Most reports have overlooked a crucial detail in the prospectus: SpaceX plans to integrate orbital computing satellites into its Starlink network roadmap by 2028.

Orbital computing satellites, as the name suggests, carry AI computing nodes onboard, forming a distributed computing network in space.

If this roadmap materializes, future AI computing power will reside not only in ground data centers but also in Earth's orbit. The implications for the global computing landscape far outweigh the significance of a successful Starship launch.

Meanwhile, China is accelerating its low Earth orbit satellite deployment (layout). The pace of Qianfan constellation deployments, Huawei HiSilicon's satellite communication chips, and a surge of commercial space startups are all vying for a foothold in this arena.

Could orbital computing become the next battleground in the U.S.-China AI rivalry?

Musk excels at pitching an "unproven" business narrative to capital willing to bet on the future at the opportune moment. SpaceX's prospectus is another testament to this methodology.

This article is intended solely for information sharing and industry analysis and does not constitute investment advice, investment analysis, or a solicitation to trade. Markets carry inherent risks; invest cautiously. Anyone making investment decisions based on this article assumes all risks and outcomes independently—the author and publishing platform bear no legal responsibility.

This article is an original work by BT Finance and may not be used, reproduced, disseminated, or adapted without permission. Infringement will result in legal action.