Anthropic revenue run rate crosses $30 billion as AI demand surges; seals partnerships with Broadcom, Google

Artificial intelligence firm Anthropic has reported a sharp surge in growth, with its annualised revenue run rate crossing $30 billion, up from $9 billion at the end of last year -- underscoring the breakneck pace of enterprise adoption of generative AI.

The company also confirmed deeper strategic partnerships with Broadcom and Google to support its rapidly expanding computing needs, as demand for its Claude AI services accelerates globally.

Enterprise demand fuels growth

Anthropic said usage of its Claude models has surged this year, with more than 1,000 enterprise customers now spending over $1 million annually -- more than double the figure recorded as recently as February.

The revenue run rate metric, commonly used by fast-growing technology firms, extrapolates current sales over a full year, offering a snapshot of momentum rather than actual booked revenue.

Chief Financial Officer Krishna Rao said the partnerships would help the company build "the capacity necessary to serve the remarkable growth" in its customer base.

Strategic chip and compute tie-ups

The collaboration with Broadcom and Google -- first announced last month -- is central to Anthropic's scaling strategy.

Broadcom is developing custom chips based on Google's tensor processing units (TPUs), offering an alternative to graphics processing units from rivals such as Nvidia. These chips are designed to handle the intensive workloads required for training and deploying large AI models.

Under a long-term agreement, Broadcom and Google will provide both chip supply and capacity assurance through 2031.

The three companies are also expanding their strategic collaboration to give Anthropic access to around 3.5 gigawatts of computing power starting in 2027 -- a massive scale that highlights the infrastructure race underpinning the AI boom.

Broadcom said the extent of Anthropic's consumption of this capacity would depend on its continued commercial success, with discussions ongoing with financial and operational partners to support deployment.

Legal overhang with US government

The growth momentum comes despite a high-profile dispute with the US government. Anthropic is challenging the Pentagon's decision to label it a supply-chain risk following disagreements over AI safety guardrails.

The company has warned that the designation could cost it billions in lost revenue. During recent court proceedings in San Francisco, its legal counsel said more than 100 enterprise clients had reached out expressing concerns about continuing engagements.

However, Anthropic's Chief Commercial Officer Paul Smith indicated that some customers view the company's stance as a reflection of its principles, potentially strengthening trust among certain segments.

AI arms race intensifies

The developments come amid intensifying competition in the AI infrastructure space, where chipmakers and cloud providers are racing to meet soaring demand.

Broadcom shares rose following the announcement, with CEO Hock Tan previously indicating that the company expects AI chip revenue to exceed $100 billion next year -- positioning it as a stronger challenger to Nvidia's dominance.

Google's TPUs, originally built to enhance its core search operations, have evolved into a critical component of the AI ecosystem. Broadcom, in turn, translates these designs into manufacturable chips, enabling large-scale deployment.