Anthropic's IPO prospectus, seen by Reuters, shows 2025 revenue of $4.59B, up 1,088% from $386M in 2024, and a GAAP net loss that widened to $41.97B from $8.31B. Roughly $34B of the loss was a non-cash accounting charge tied to the rising value of financing that could convert into Anthropic shares, not money spent running the business. The operating loss expanded to $8.06B from $2.98B as compute and infrastructure spending rose 190% to $7.33B, or 58% of its $12.65B in total operating expenses and more than its revenue for all of 2025.

Sales have since surged, with preliminary second quarter revenue above $11.5B, more than double all of 2025, and adjusted operating income positive. Annual recurring revenue reached about $65B by late July. The prospectus outlines $518B in cloud, compute and infrastructure obligations in the coming years, a total Reuters initially reported as due within one year before correcting its story. Anthropic ended 2025 with $20.28B in cash, cash equivalents and short-term investments. Its two largest direct customers accounted for 12% of revenue each, and the company warned that many of its largest clients are not locked into long-term contracts and could reduce or stop spending.

The IPO could value Anthropic at more than $2T, more than double the $965B post-money valuation of its May funding round, in what would be the biggest IPO in history. The listing could come after the November US midterm elections. Risk factors occupy roughly 80 of the filing's 261 pages, against 48 describing the business, including the warning that advanced AI could pose catastrophic or existential risks to humanity. Amazon and Google, both major investors in Anthropic, supply much of the cloud infrastructure used to train and run Claude.