Anthropic Will Beat OpenAI to Public Markets. That Sequence Sets Every AI Valuation That Follows.

Two of the most valuable private companies in history are filing for IPO within two weeks of each other. When they land on public exchanges — almost certainly before the end of 2026 — they will establish the pricing reference for every AI company that follows. The sequencing and the multiples will define an asset class in real time.

Anthropic filed confidential S-1 paperwork with the SEC on June 1, TechCrunch confirmed. OpenAI had filed its own S-1 on May 22, targeting a $1 trillion valuation on $25 billion in annualized revenue at a 33% gross margin. Anthropic's metrics at filing: $47 billion in annualized revenue, projected 130% growth to approximately $108 billion next year, first operating profit approaching. The financial profiles are meaningfully distinct — and that distinction will shape how public market investors price both companies simultaneously.

OpenAI's public market challenge is structural. At 33% gross margin, every dollar of revenue costs 67 cents in compute before a single employee is paid. HSBC's modeling projects a $207 billion capital requirement through 2030 to honor existing compute commitments. The path to the $1 trillion valuation requires either dramatic compute cost compression — which is happening, but incrementally — or a product mix shift toward higher-margin enterprise software. Both require time that public market investors will price with discipline that private capital did not.

Anthropic's profile is more favorable. The company attributes its revenue growth to enterprise and developer API consumption — production deployments, not consumer subscription. The investor list in the $65 billion Series H — Altimeter Capital, Dragoneer, Capital Group, Coatue, D1 — is composed almost entirely of crossover specialists who anchor public offerings and hold positions through lock-up. Those checks were pre-positioning for a public exit they believe is on a specific, visible timeline. Altimeter doesn't write $965 billion pre-money checks as a decade-long bet.

The race has consequences for every other AI company in the market. When Anthropic and OpenAI trade publicly, they establish the multiples against which all downstream AI investment will be benchmarked. Application-layer companies will be compared to foundation model companies for the first time in a liquid market. If foundation models trade at 15x revenue and application-layer AI trades at 10x, the allocation decision between them becomes explicit in a way private market pricing never forced. The implicit question — "is this company worth foundation model multiples or application multiples?" — becomes a daily pricing question.

The detail worth watching isn't the IPO valuation at listing. It's what the public market does with both companies in the twelve months after. Foundation model companies going public at sub-50% gross margins need to demonstrate a margin expansion path. The application-layer companies whose economics were always structurally better — higher margins, lower compute dependency, deeper vertical data moats — will look increasingly reasonable against a public market that has now priced the foundation layer with visible constraints. The IPO isn't the endpoint of the AI investment story. It's the moment the market starts pricing it with rigor.

Anthropic beating OpenAI to market by a few weeks is a footnote. What isn't a footnote is what both filings collectively tell the market about where AI value accumulates — and whether public investors agree with the answer private capital has been paying for since 2023.