Fortune Tech: What Anthropic’s IPO prospectus reveals

Anthropic has finally opened the books, and the picture is as strange as the filing’s apocalyptic flourishes. The company filed for its public offering on Monday, kicking off what could be one of the biggest IPOs ever and setting up a Nasdaq listing under the ticker ANT, likely around Thanksgiving week.

The headline number is wild: revenue jumped 1,300% in the first half of the year to $16.3 billion, powered by demand for the Claude family of AI products. But the same filing also shows how expensive this kind of growth is. Anthropic spent $5.6 billion on compute in the second quarter of 2026 alone, even as it said that every dollar of compute produced $2.08 in revenue.

The loss figures are brutal at first glance. Anthropic reported a net loss of $80.7 billion for the six months through June 2026, though $72.4 billion of that came from an accounting charge tied to convertible notes. On an operating basis, the picture is less theatrical: losses narrowed to $693 million, down from $2.1 billion in the same period of 2025. The company also said it was profitable on an adjusted basis in the first half, with $1.6 billion in operating income.

The filing also puts a spotlight on how tightly Anthropic is controlled. The San Francisco company was founded in 2021, had 5,300 employees at the end of September, and its seven cofounders hold the power thanks to a multi-class share setup. In the letter at the top of the prospectus, they pitched the company as a guide for a safer technological future, while the risk section warned that losses will keep coming as the business keeps spending heavily and adds the costs of being public.