Three weeks ago, the story was OpenAI's audited losses. The picture has moved considerably since then.
SpaceX went public on June 12, priced at $135 per share. With roughly 13.1 billion shares outstanding, that put its opening market value at $1.77 trillion, the largest IPO on record. The stock closed at $161, a 19% gain on day one. The company is now the sixth-largest publicly traded business in the United States by market value.
The SpaceX debut matters for OpenAI and Anthropic because it was, in the words of analysts at Capital Economics, a test of whether investors will fund companies that combine enormous ambitions with negative cash flow. The test passed. Retail demand was so intense that individual investors on platforms like Robinhood received only a fraction of the shares they requested. The broader market for large AI-era IPOs is open, at least for now.
OpenAI filed its own confidential IPO paperwork on June 8, one week after Anthropic did the same. OpenAI is targeting a listing as early as September. Its annualized revenue run rate sits at roughly $25 billion, up from $13 billion in full-year 2025. The company expects to lose around $14 billion in 2026 even as revenue grows, with profitability not projected until 2030. Its CFO has reportedly raised concerns internally about whether the company can sustain its data center spending commitments.
Anthropichurried into a very different financial position. Annualized revenue crossed $47 billion in May 2026, up from $9 billion at the end of 2025. That is a pace of growth that few companies in history have achieved. More striking: Anthropic told investors it projects a $559 million operating profit for the quarter ending June 2026, on revenue of $10.9 billion. That would be the first profitable quarter for any frontier AI lab. Anthropic itself warned investors that profitability may not hold, because planned data center spending will increase costs later in 2026. But the direction is clear, and it matters.
There is now a genuine gap between the two companies going to market. Anthropic's annualized revenue leads OpenAI's. Anthropic's path to profitability is shorter: it is targeting positive free cash flow by 2027, three years before OpenAI's own estimate. Anthropic is also more dependent on business customers, with around 80% of revenue coming from companies rather than individual subscribers. That gives it a more predictable income base and less exposure to consumer churn.
The SpaceX IPO itself surfaced one important data point that shapes the whole picture. In its filed documents, SpaceX disclosed that Anthropic is paying $1.25 billion per month for access to computing capacity in SpaceX's data centers, through 2029. That is $15 billion a year in committed spending, which explains both why Anthropic's costs are high and why SpaceX's valuation story includes AI infrastructure as a core business line.
For the companies and professionals who use these tools, the IPO race has a practical consequence. Once OpenAI and Anthropic are public, they will report quarterly results and face pressure to show margin improvement. The most direct outcome is that pricing for AI tools will rise over time, and the tolerance for subsidising users at below-cost rates will shrink. Forrester research found that 25% of planned enterprise AI spending is already being delayed into 2027 as finance teams begin asking harder questions about what they are getting for their money. That spending caution is a risk to every revenue projection these companies are showing investors.
Perplexity's CEO said publicly that the SpaceX IPO would be a leading indicator for how OpenAI and Anthropic are received. SpaceX passed that test. The harder one comes when investors see the full financial details, which will be published before each company lists. Those documents will show exactly what it costs to keep these services running and how much of every dollar customers spend actually stays in the company. The answers will set the terms for every pricing, contract, and technology decision that follows.