On April 6, 2026, AI safety and enterprise large‑model leader Anthropic announced its annualized revenue (Annualized Revenue Run‑rate, ARR) had surpassed $30 billion. That figure is more than three times the $9 billion recorded at the end of 2025 and is a substantial rise from $14 billion in February 2026, making it one of the AI industry’s most striking growth stories. At the same time, Anthropic is jointly undertaking a major compute expansion with Google and Broadcom, expected to add more than 1 gigawatt (GW) of AI compute capacity in 2026, with investments totaling tens of billions of dollars.
Anthropic CEO Dario Amodei said in internal meetings and recent earnings communications that the surge is primarily driven by the Claude product line, especially Claude Code. As a developer‑oriented AI coding tool, Claude Code’s annualized revenue has exceeded $2.5 billion, roughly doubling since early 2026, while weekly active users have also doubled in the same period. Enterprise subscription revenue accounts for over 50% of total revenue, and the number of large corporate clients has increased almost sevenfold year on year. Anthropic highlights that, in under three years since founding, the company has moved from its first dollar of revenue to a $30 billion ARR, with average annual growth exceeding tenfold over the past three years.
Business logic behind the revenue surge
Anthropic’s revenue profile is highly enterprise‑centric, in sharp contrast to OpenAI’s more consumer‑oriented approach. Enterprise customers contribute the bulk of revenue, and Claude Code already represents about 4% of public GitHub submissions and continues to rise. This trend reflects AI’s shift from a “chat tool” to foundational productivity infrastructure, with enterprises willing to pay premium prices for reliable, secure, and controllable AI coding and agent capabilities.
However, rapid growth also entails heavy investment. Anthropic remains deeply loss‑making, largely due to elevated compute costs. To meet the training and inference demands of Claude models, the company must continuously scale its infrastructure. Crossing the $30 billion ARR threshold strengthens Anthropic’s financing and cash‑flow position and has further bolstered market confidence in its valuation (its prior Series G placed a post‑money valuation at $380 billion).
Partnering with Google and Broadcom: a multibillion‑dollar compute arms race
Alongside rapid revenue growth, Anthropic is making an aggressive push into AI infrastructure through deep partnerships with Google and Broadcom:
- Google Cloud TPU collaboration: Anthropic plans to procure up to one million units of Google’s next‑generation TPUs, expected to add more than 1 GW of compute capacity in 2026, with investments running into the tens of billions of dollars. Google will provide customized cloud services and data center support, marking the latest upgrade in a long‑term partnership.
- Broadcom custom chips: Broadcom has confirmed Anthropic as a major customer, involving multiple orders each worth billions of dollars. Broadcom will deliver complete server racks that incorporate TPUs and other components to help Anthropic rapidly stand up high‑performance computing clusters. Broadcom CEO Hock Tan stated on an earnings call that Anthropic is the company’s fourth‑largest customer for its XPU (custom AI chips) business, and order volumes continue to grow.
In addition, Anthropic is making a separate $50 billion investment to build data centers in the United States (including Texas and New York), with some projects co‑financed with Google. Several sites are expected to come online in 2026, creating thousands of jobs. These facilities will serve next‑generation Claude model training and enterprise deployments, ensuring the company remains competitive with rivals such as OpenAI and Google.
Strategic significance and industry impact
Anthropic’s dual focus on “revenue + compute” highlights a broader industry shift from a pure “model race” to a balance of infrastructure investment and commercialization. A $30 billion ARR not only validates the massive market potential of enterprise AI but also lays a foundation for a potential IPO or further financing. The joint compute expansion with Google and Broadcom is a proactive response to global GPU/TPU supply constraints and aims to secure sustained leadership in training and inference capacity.
For the wider industry, this development accelerates the AI compute arms race. Google’s commitment of TPUs and data center resources to Anthropic further cements its cloud market position, while Broadcom benefits from growing custom‑chip orders. Chinese large‑language model vendors (such as Tongyi Qianwen and DeepSeek) may retain advantages in cost‑performance, but the gap in infrastructure investment could widen. The global AI supply chain—including chips, data centers, and energy—looks set for a new wave of investment.
Challenges and outlook
Despite the surge in revenue and compute, Anthropic still faces high losses, intense talent competition, and regulatory pressures (including the EU AI Act and U.S. security reviews). The company emphasizes a “safety first” approach and plans to continue strengthening alignment techniques such as constitutional AI while expanding.
Anthropic CFO Krishna Rao said, “We are at a critical juncture as AI moves from experimentation to widespread adoption. A $30 billion ARR and large‑scale compute expansion will help us bring reliable, beneficial AI to enterprises and users around the world faster.” With next‑generation Claude models and infrastructure coming online, Anthropic could reach positive cash flow in 2027–2028 and emerge as a core player in AI infrastructure.
From a $30 billion ARR to multiple gigawatts of compute, Anthropic is rapidly scripting an enterprise‑AI success story. Its dual “revenue + compute” growth model offers a replicable template: in the AI era, the ability to monetize commercially and to foresee and build infrastructure are decisive factors.
▶️ Like this article? Creating this content is hard!
Subscribe to our YouTube channel:
AI Rosetta
Your likes, subscriptions and shares are our greatest support! 🙏
For more AI news visit: https://www.airosetta.com