Analysis
Anthropic's preliminary second-quarter revenue exceeded $11.5 billion, according to Bloomberg and confirmed by CNBC.
That's up more than 14-fold from a year earlier, and a sharp acceleration from its own first quarter. The company also posted positive adjusted operating income for the first time, a milestone that arrives just as it prepares a possible fall IPO.
Background: from $65B round to IPO track
Anthropic filed confidentially for an IPO on June 1, weeks after closing a $65 billion funding round in late May. Pulse has tracked the run-up as the company has simultaneously built out its own chip-design ambitions and continued scaling Claude enterprise deployments -- both capital-intensive bets that this revenue trajectory now appears to be funding organically rather than purely through fresh equity.
Competitive context
The obvious comparison is OpenAI, whose revenue run-rate recently topped $40 billion on an annualized basis -- roughly double Anthropic's quarterly figure annualized, but OpenAI is reportedly still projecting a $14 billion loss for 2026 even at that scale. Anthropic reaching positive adjusted operating income at roughly a quarter of OpenAI's revenue is a meaningfully different capital-efficiency story, and it's the kind of detail public-market investors will scrutinize hardest once both companies file real S-1s. Google DeepMind and Meta's AI unit don't break out comparable standalone figures, making Anthropic and OpenAI the two clearest data points for how big frontier-lab businesses can get.
The numbers in context
A 14x year-over-year growth rate at an $11.5 billion quarterly run-rate is a scale of growth almost unprecedented outside of the AI category -- for comparison, Databricks, widely seen as one of the best-run infrastructure companies of this cycle, posted "just" 80%-plus year-over-year growth on its own $7 billion revenue run-rate disclosed this month. Anthropic's growth rate at a larger revenue base suggests enterprise Claude adoption, not just consumer subscriptions, is doing meaningful work.
What to watch next
The preliminary numbers still need to survive audit before any S-1 becomes public, and Anthropic's own risk report this month noted the company has no current plans to release a more powerful "Model 2" despite rising AI-risk concerns -- a signal the company may be deliberately pacing capability releases ahead of its public listing to avoid inviting fresh regulatory scrutiny mid-roadshow. If the October IPO timeline holds, this quarter's numbers become the baseline the market prices against.