Anthropic's Revenue Run Rate Clears $65B as a $2 Trillion Listing Comes Into View

Anthropic's annualized revenue run rate passed $65 billion at the end of July, according to Bloomberg — a figure that would have looked implausible for any software company a year ago, and one the model maker reached from a standing start of roughly $9 billion at the close of last year.

The trajectory matters more than the number. The run rate, which projects a full year from a recent shorter period, sat at $47 billion in May. Adding $18 billion of annualized revenue across two months means growth is not merely fast but still accelerating, at a scale where deceleration is usually the default. Anthropic did not immediately respond to a request for comment on the figures.

What investors are modelling next

Backers expect the pace to hold through the rest of the year, finishing 2026 somewhere between $100 billion and $120 billion, the Financial Times reported. That projection is doing heavy lifting in how the company is being valued, because it implies the enterprise and developer demand driving the last two quarters is structural rather than a burst of early adoption.

Rival OpenAI, by comparison, has doubled its revenue to $40 billion from $20 billion at the end of 2025, Bloomberg reported last week. The two firms may compute their metrics differently, so a direct comparison deserves caution. Still, the gap in growth rate — not absolute size — is what has captured investor attention, and Anthropic's curve is the steeper one.

An IPO ahead of the field

Both companies have filed confidential IPO paperwork, and Anthropic is expected to reach the public markets first, potentially as soon as this autumn. The FT reported it will seek a public valuation of $2 trillion or more, which would make it the largest market debut on record by a wide margin.

For context on how quickly that bar has moved: Anthropic was last valued at $965 billion in late May, when it closed a $65 billion round. A $2 trillion listing would roughly double a private mark set only three months earlier.

Why the numbers are being scrutinised

Annualized run rate is a forward projection, not booked revenue, and it flatters any business whose most recent month is its best. That is precisely why the metric has become the industry's shorthand during the current buildout, and why public-market investors are likely to press for cleaner disclosure once a prospectus lands.

The other open question is cost. Frontier model training and inference consume capital at a rate that revenue growth alone does not settle, and the AI leaders have so far disclosed little about the relationship between the two. A listing would force some of that into the open.

Outlook

If Anthropic does price this autumn at anything near the reported target, it will set the reference valuation for every AI company that follows, including OpenAI. A strong debut would validate the argument that model providers are infrastructure businesses with durable enterprise revenue. A soft one would reframe the same revenue curves as a bet placed at the top of a cycle.

For now the sequencing favours Anthropic. Being first to list means setting the narrative rather than inheriting it, and the growth figures reported this week are the strongest evidence yet that the company intends to be measured on trajectory rather than scale.