Anthropic S-1 shows $16.3B revenue and $80.7B loss ahead of Nasdaq listing as ANT

Anthropic has taken its most concrete step yet toward the public markets. According to a summary published on dev.to, the company behind the Claude AI models filed a public S-1 registration on October 26, with plans to list Class A shares on the Nasdaq Global Select Market under the ticker ANT. The filing describes a business scaling at an extraordinary rate, alongside a headline net loss that is largely an accounting effect rather than a cash problem.

Revenue is compounding at unusual speed

The figures dev.to pulls from the document show revenue of $16.286 billion for the first half of 2026, roughly fourteen times the $1.163 billion recorded over the same stretch a year earlier. Full-year 2025 revenue came in at $4.585 billion, and the company reportedly generated just $41 million back in 2023. The customer base is deepening too: more than 11,900 customers now spend at least $100,000 a year with Anthropic, and dev.to reports that existing customers are laying out roughly nine times what they spent a year ago.

The $80.7 billion loss is mostly an accounting charge

The number that will dominate coverage is a net loss of $80.748 billion. But according to dev.to, $72.4 billion of that is a non-cash accounting charge connected to convertible notes and related instruments. The underlying operations tell a very different story: an operating loss of $693 million for the period, a GAAP operating profit of $974 million in the second quarter, and positive operating cash flow of $3.929 billion. On that reading, the business itself sits close to breakeven, and the enormous net-loss line is mostly paper.

Key deal terms are still blank

Several essentials are missing from the initial document. The share count, price range, and listing date are all left empty and will arrive in an amended filing. The registration is not yet effective, so no securities can actually be sold, and the roadshow cannot begin before mid-November at the earliest. Goldman Sachs and Morgan Stanley are leading a syndicate of roughly twenty banks. On governance, a founder-controlled LLC retains 50.1% of the voting power, meaning public shareholders would hold economic exposure to Anthropic without a say in its direction.

The compute bill behind the raise

The filing also discloses commitments of up to $807 billion for data center and chip contracts over the next ten years, much of it non-cancellable. In dev.to's assessment, this is what the IPO proceeds are fundamentally for: locking in compute capacity for the decade ahead. As for the valuation figures between $1.5 trillion and $2 trillion circulating in the press, dev.to notes that they do not appear anywhere in the filing itself.

Why it matters

This is, in dev.to's characterization, the biggest IPO story in tech at the moment, and it offers the clearest public look yet at the finances of a frontier AI lab. Two numbers will frame the investor conversation: revenue growing roughly fourteenfold year over year, and an $807 billion compute bill that must be paid whether or not that growth continues. The structure of the deal matters just as much. Founders keep voting control, so public buyers are effectively funding infrastructure commitments rather than acquiring influence over the company. For anyone following the AI industry, the S-1 turns abstract debates about scaling, capital spending, and concentrated power into concrete figures that can now be examined and argued over.