Anthropic's annual run rate rose from $1 billion to $19 billion in 14 months
The $19 billion figure was already out of date, and the two prior years had each brought tenfold growth.
Lenny's Podcast: Product | Career | Growth · Head of Growth (Anthropic): “Claude is growing itself at this point” | Amol AvasareAnthropic, the artificial intelligence company behind Claude, rose from $1 billion in annualized recurring revenue at the start of 2025 to $19 billion about 14 months later. Annualized recurring revenue is a run rate, a projection of a full year from recent results.
Figures cited in the discussion put the climb at about $4 billion in mid-2025 and about $9 billion at the end of 2025. Amol Avasare, Anthropic's head of growth, put the earlier path at zero to $100 million in 2023, $100 million to $1 billion in 2024, and about $1 billion to about $10 billion in 2025. Only the second and third stretches are tenfold increases, because the 2023 run started from nothing.
Avasare called the rise "a complete miracle" for a company that began as the smallest and least funded rival. He said the $19 billion figure was from late February and was already out of date, and that growth had not slowed. He joked that straight line charts are not cool inside the company, where everything is shown on a logarithmic scale.
Questions people ask
What does ARR mean here?
ARR stands for annualized recurring revenue. It takes subscription and usage revenue from one month or quarter and projects what it would total over a full year. For fast-growing software and AI companies, it is a standard way to show current scale.
How does Anthropic make that much money?
It sells access to its Claude models to businesses and developers, plus paid subscriptions for individuals and teams. Its best-known strength is writing and editing computer code, along with business and office work. Companies pay for use through an interface and through products such as Claude Code.
Is Anthropic profitable at $19 billion in revenue?
Revenue alone does not mean profit. Training and running large AI models requires enormous spending on chips, data centers, and electricity. Anthropic has raised tens of billions of dollars, and published reports have described large ongoing losses alongside the revenue growth.