Anthropic's Revenue Triples to $30 Billion Run Rate in Three Months — Signs Largest Compute Deal Yet with Google and Broadcom

When a company's revenue triples in a single quarter, the infrastructure bill arrives immediately. Anthropic announced this week that its annualized revenue has surpassed $30 billion — up from $9 billion at the end of 2025 — and simultaneously signed the largest compute deal in its history: 3.5 gigawatts of Google TPU capacity through Broadcom, coming online in 2027. The numbers represent the fastest revenue acceleration any AI company has achieved, outpacing OpenAI's estimated $25 billion annualized run rate.

What Is Driving the Revenue Surge?

The $9 billion to $30 billion jump happened in roughly three months. Enterprise customers spending $1 million or more per year doubled from over 500 in February 2026 to more than 1,000 now, according to Anthropic's disclosures. That doubling took under two months.

Several factors explain the pace. Claude Code, Anthropic's developer tooling product, has driven adoption among engineering teams at companies that previously split spend across multiple AI providers. Claude is the only frontier AI model available on all three major cloud platforms — AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry — which removes a procurement barrier that competitors still face. An enterprise buyer on any cloud can access Claude without switching infrastructure.

The company's February 2026 Series G round raised $30 billion at a $380 billion valuation, making it one of the most valuable private companies in the world. At the time, some analysts questioned whether the valuation was justified. Three months later, the revenue trajectory suggests the market priced in growth that has now materialized.

How Big Is the Compute Deal?

Anthropic CFO Krishna Rao called the new agreement "the most significant compute commitment to date." The deal secures 3.5 gigawatts of next-generation Google TPU capacity, manufactured by Broadcom, with delivery starting in 2027. It extends an existing arrangement from October 2025 for over 1 gigawatt of TPU capacity that is already being supplied in 2026.

The majority of new compute will be sited in the United States, extending Anthropic's $50 billion domestic infrastructure commitment announced in November 2025. Broadcom CEO Hock Tan said in an SEC filing that "for Anthropic, we are off to a very good start in 2026." Mizuho analysts estimate Broadcom's AI revenue from Anthropic alone will reach $21 billion in 2026 and $42 billion in 2027 — numbers that make Anthropic one of Broadcom's single largest customers.

Anthropic trains Claude across a diversified compute stack: AWS Trainium chips, Google TPUs, and NVIDIA GPUs. The multi-provider approach reduces dependency on any single hardware supplier and gives the company leverage in pricing negotiations that smaller AI labs cannot match.

Why Is Revenue Growing Despite the Pentagon Conflict?

The growth is happening against an unusual political backdrop. The US Department of Defense labeled Anthropic a "supply chain risk" after the company declined to deploy Claude on classified DoD networks. That same classified-AI contract triggered the #QuitGPT movement when OpenAI accepted it — employees resigned over concerns about military applications of AI models.

Anthropic chose the opposite path and absorbed the Pentagon's designation. Despite the label, its enterprise customer count doubled and revenue tripled. The commercial market appears to be rewarding the safety-first position rather than penalizing it. Companies spending $1 million or more per year on Claude are choosing the model on capability and availability, not on its relationship with the defense establishment.

The company is also expanding internationally. On April 1, Anthropic signed a Memorandum of Understanding with the Australian government and opened a Sydney office — its first presence in the Asia-Pacific region. The deal includes AI safety research collaboration, university funding, and startup API credits. It is part of a pattern: Anthropic is building government relationships in allied democracies while maintaining distance from classified military applications.

Mubboo's Take

When we reported on Anthropic's Australian MOU two days ago, we noted it was significant for local AI startups building on Claude. This week's revenue disclosure puts that deal in a much larger context. Anthropic is not a niche safety-focused lab making principled but commercially costly decisions — it is now a $30 billion revenue company growing faster than any competitor, with infrastructure commitments secured years in advance. For consumer platforms that build on Claude, the practical implication is straightforward: the model we depend on is backed by the fastest-growing AI company in the world, with 4.5 gigawatts of compute capacity either online or under contract. That is the kind of infrastructure stability that matters when you are building across multiple markets.