Anthropic's Revenue Soars Past $30B, Inks Major Chip Deals with Broadcom and Google
By Bloomberg News
Anthropic PBC, the artificial intelligence startup behind the Claude chatbot, announced on Tuesday that its annual revenue run rate has skyrocketed past the $30 billion mark. This represents a staggering leap from the $9 billion run rate reported at the end of 2025, underscoring breakneck commercial adoption.
The company confirmed and detailed its previously announced strategic collaboration with chip designer Broadcom Inc. and tech giant Google, a move aimed at securing the immense computing capacity required to sustain its explosive growth. Anthropic revealed it now counts over 1,000 business customers each spending more than $1 million annually on its services, a figure that has more than doubled since February.
"Our partnerships with Broadcom and Google are about building the foundational capacity to serve the remarkable, accelerating demand from our enterprise customer base," said Anthropic CFO Krishna Rao in a statement. The growth comes even as the company navigates a high-profile legal dispute with the U.S. Department of Defense, which labeled it a supply-chain risk following disagreements over AI safety protocols.
Under the multi-faceted deal, Broadcom will develop custom AI chips using Google's Tensor Processing Unit (TPU) architecture, creating a potent alternative to the dominant technology from Nvidia Corp. The companies have entered a long-term supply agreement extending through 2031. Furthermore, the trio is expanding a collaboration to grant Anthropic access to approximately 3.5 gigawatts of computing power, with deployment scheduled to begin in 2027.
"The consumption of this expanded AI compute capacity is contingent on Anthropic's continued commercial success," Broadcom noted in a regulatory filing. The announcement buoyed Broadcom shares, which rose as much as 5% during Tuesday's trading session.
Analysts view the deals as a critical step for Anthropic to vertically integrate its hardware supply chain, reducing dependency on any single vendor. "This isn't just a procurement agreement; it's a strategic moat," said tech industry analyst Michael Chen. "Anthropic is locking in the silicon and energy it needs to compete at scale for the next decade, while Broadcom gets a flagship customer to challenge Nvidia's hegemony."
The company has warned that the Pentagon's "supply-chain risk" designation could potentially cost billions in lost revenue. However, Anthropic's Chief Commercial Officer, Paul Smith, suggested that some clients respect the firm's principled stance. "The dispute, while challenging, has also resonated with enterprises that value responsible AI development," Smith noted in a recent interview.
User Reactions:
Sarah Lin, Tech Investor: "This run rate is phenomenal and validates the enterprise AI market's depth. The Broadcom-Google partnership is a masterstroke—it mitigates supply risk and could drive down inference costs long-term."
David Park, AI Ethics Researcher: "The growth is impressive, but it amplifies the core dilemma. A company powerful enough to command 3.5 gigawatts of power is the same one in a standoff with regulators over safety. We're centralizing immense capability without resolving the governance debate."
Marcus Johnson, Startup CTO: "It's getting impossible for smaller players to compete. When you need deals measured in gigawatts and billions just to have a seat at the table, it stifles innovation. This is consolidation masquerading as partnership."
Priya Mehta, Enterprise Procurement Lead: "The stability offered by a 2031 supply pact is exactly what large corporations want to see before making billion-dollar AI commitments. It turns Claude from a cutting-edge tool into a viable long-term platform."