Anthropic has filed its S-1 registration statement with the US Securities and Exchange Commission (SEC) as it advances towards an initial public offering (IPO) at an implied valuation, reportedly between $1.5tn and $2tn.
The Claude developer, which is expected to debut on the stock market before Thanksgiving, disclosed rapid revenue growth alongside substantial losses in the filing.
Revenue in the three months ended 30 June surged 1,368% compared with the same period a year earlier, although the IPO prospectus cautioned that the pace of growth could decline and fluctuate over time.
Anthropic’s net loss for the first half of 2026 widened sharply to $80.7bn, a figure driven primarily by a non-operating charge tied to bonds convertible into the company’s equity.
On an operational basis the trajectory was considerably stronger. The operating loss for the first half of the year narrowed to $693m, while adjusted operating income turned positive at $1.62bn, reversing a loss of $1.85bn recorded in the corresponding period of 2025.
As of 30 September, Anthropic counted more than 11,900 enterprise customers generating at least $100,000 in trailing-12-month revenue, with more than 1,900 exceeding the $1m threshold and roughly 250 surpassing $10m.
Net revenue retention reached approximately 900% at the same date, and trailing-12-month revenue per employee stood above $6.6m.
Anthropic has also amassed significant financial resources to support its capital-intensive operations.
By 30 June 2026, the company had raised a cumulative $141.1bn and held $112.4bn in cash and equivalents alongside $19.3bn in short-term investments.
The filing explicitly detailed a series of massive multi-billion-dollar infrastructure and cloud commitments. In the prospectus, Anthropic disclosed specific long-term minimum and maximum agreements with partners including Broadcom, SpaceX, Google, and Amazon to secure the compute power necessary for its next-generation models.
Established in 2021, Anthropic characterises itself as a research laboratory building advanced AI systems with a focus on long-term safety and societal benefit.
Its flagship product, Claude, is delivered through a suite of models, a developer platform and end-user applications, all oriented towards enterprise customers.
Anthropic frames Claude as a “Virtual Collaborator” designed to execute complex, multistep tasks in tandem with human users.
Two specific offerings highlight this vision: Claude Code, which plans, writes, tests and debugs software, and Claude Cowork, which brings comparable autonomous capabilities to a wider range of knowledge work.
The prospectus also outlined risks, warning that AI misuse, cyberattacks, the concentration of power and the behaviour of misaligned autonomous systems could each cause serious harm.
The filing arrives at a moment when investors are closely evaluating whether the scale of capital flowing into AI infrastructure will ultimately produce commensurate returns.
