Anthropic's rapid revenue surge

Anthropic says its annual revenue run rate has passed $30 billion, up from about $9 billion at the end of 2025, which means its recent monthly sales have more than tripled in roughly one quarter. The company disclosed the figure on April 6 while announcing a new computing deal with Google and Broadcom. (anthropic.com) That number is a run rate, not booked revenue for a full year. It is the startup version of taking one very fast lap and asking what the full race would look like if the pace holds. (economictimes.indiatimes.com) Anthropic also said more than 1,000 business customers are now spending over $1 million a year on Claude, up from more than 500 when it announced its Series G fundraising in February 2026. That means the growth is not coming from a handful of giant contracts alone. (anthropic.com 1) (anthropic.com 2) The company has been building for this exact market. In March, Anthropic launched a $100 million Claude Partner Network to train consulting firms and software integrators that help large companies move from pilot projects to full deployments. (anthropic.com) It has also been leaning hard into the kind of features big companies buy first: coding help, long context windows, complex agents, and tighter cloud integrations. Anthropic’s flagship Claude models are sold directly through its own platform and through Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry. (anthropic.com 1) (anthropic.com 2) The other half of this story is electricity-sized computing demand. Anthropic said its new agreement with Google and Broadcom will add multiple gigawatts of next-generation Tensor Processing Unit capacity starting in 2027, with most of that infrastructure located in the United States. (anthropic.com) A gigawatt is power-plant scale. When an artificial intelligence company talks about multiple gigawatts, it is saying the bottleneck is no longer just software talent or model design, but how many chips and how much power it can lock up before rivals do. (anthropic.com) At the same time, Anthropic is giving selected partners early access to an unreleased model called Mythos Preview, but not as a general product launch. The company says it will not make Mythos generally available and is instead using it inside Project Glasswing, a defensive security effort focused on finding and fixing dangerous software flaws. (anthropic.com) (securityweek.com) Anthropic says more than 40 organizations that build or maintain critical software infrastructure are getting access to Mythos Preview to scan first-party and open-source systems. That turns a frontier model into a wedge for deeper enterprise relationships, because the customers testing security tools today are often the same ones choosing long-term model providers tomorrow. (anthropic.com) So the picture is not just “Anthropic got bigger.” It is that one company is simultaneously stacking revenue, million-dollar customers, consulting partners, cloud distribution, and future chip supply, which is how software platforms stop looking like experiments and start looking hard to dislodge. (anthropic.com 1) (anthropic.com 2)