Anthropic files for IPO at $965B, eclipsing OpenAI in valuation
Anthropic confirmed on Monday that it had quietly handed the SEC a draft Form S-1, joining OpenAI on the runway to a public listing and ending any pretense that the leading AI labs would stay private through the current funding cycle. The announcement landed ten days after OpenAI's own confidential filing on May 22, but it carries the number that matters to anyone underwriting an AI bet right now: a $47 billion annualized revenue run rate, up from roughly $9 billion when 2026 began.
What was actually filed
The company's own statement is one paragraph long: "Anthropic, PBC confidentially submitted a draft registration statement on Form S-1 to the U.S. Securities and Exchange Commission" and "the number of shares to be offered and the price have not yet been set." A confidential draft is a JOBS Act filing that lets the company iterate on disclosures with the SEC without making them public, and converts to a public S-1 weeks before pricing. Goldman Sachs, Morgan Stanley, and JPMorgan Chase are the banks under consideration for the lead underwriting roles, per reporting shared between Anthropic and OpenAI mandates. A fall debut is on the table; an October target keeps showing up in the most concrete reporting.
The filing also closes a private-market gap that mattered. Anthropic's $65 billion Series H, which closed in late May at a $965 billion post-money valuation led by Altimeter, Dragoneer, Greenoaks, Sequoia, Capital Group, Coatue, and D1, vaulted it past OpenAI's $852 billion. That was the first time any other private AI company had out-priced OpenAI on the secondary book.
Why it matters
The growth curve is the headline finding. Anthropic told investors revenue tripled from around $4 billion at the end of 2024 to $9 billion in January, and it has roughly quintupled again since — driven, its investor materials suggest, by enterprise adoption of Claude for coding and agentic workflows rather than consumer subscriptions. The company told investors the run rate will exceed $50 billion by the end of July.
That mix matters because public markets price recurring enterprise revenue very differently from API or chatbot traffic. If the S-1 confirms the bulk of that $47 billion is coming from contracted enterprise spend on Claude Code and agent platforms, Anthropic will go out looking more like a high-growth SaaS company than an AI consumer brand, and that is the comp that justifies a near-trillion valuation. If a meaningful share turns out to be lumpy hyperscaler reselling or compute pass-through, the number will get a haircut on the roadshow.
The counter is what every AI bear has been saying out loud since the OpenAI filing two weeks ago: revenue at this scale, this fast, has only ever been possible because the underlying compute is being sold at or below cost by partners with their own strategic reasons to subsidize it. A public S-1 will force Anthropic to disclose its capacity commitments to AWS, Google, and others, and the gap between gross revenue and contribution margin is what everyone — Wall Street, the AI critics, and Anthropic's own competitors — will fight about for the rest of the year.
What to watch
The first public S-1 amendment is the one to read. That is when the confidential drafting stops and the actual numbers — gross margin, compute commitments, customer concentration, the split between API and enterprise contracts — become public for the first time. Watch for it sometime between mid-August and mid-September, three to six weeks before any proposed pricing. OpenAI's first public amendment will likely arrive on a similar clock, which makes the back half of 2026 a side-by-side disclosure event the AI sector has never had.
Sources