Author: Deep Tide TechFlow
Deep Tide Introduction: On April 6, Anthropic disclosed that its annualized revenue exceeded $30 billion, more than doubling from $9 billion at the end of 2025. The number of enterprise customers with annual spending exceeding $1 million doubled from 500 to 1000 within two months. On the same day, Broadcom's SEC filing confirmed that Anthropic will acquire approximately 3.5 gigawatts of next-generation TPU computing power starting in 2027, marking its largest single computing power commitment to date.
Anthropic released two significant figures on the same day.
According to Anthropic's official blog on April 6, the company's annualized revenue has surpassed $30 billion, more than doubling from approximately $9 billion at the end of 2025. On the same day, Broadcom's SEC filing disclosed that Anthropic will acquire about 3.5 gigawatts (GW) of next-generation TPU computing power through Broadcom starting in 2027, as part of an expanded collaboration among the three parties (Broadcom, Google, and Anthropic).
Broadcom's stock rose approximately 3% in after-hours trading.
(Image source: X user @damianplayer)
From $1 Billion to $30 Billion in 14 Months: Claude Code Is the Core Engine
Anthropic's revenue growth curve is unprecedented in the AI industry. According to publicly disclosed information and reports from Bloomberg and other media outlets, the timeline is as follows: approximately $1 billion in December 2024, about $4 billion in mid-2025, around $9 billion at the end of 2025, roughly $14 billion in February 2026, nearly $19 billion in early March, and officially confirmed to exceed $30 billion on April 6.
Anthropic CFO Krishna Rao stated in a declaration that the company is making "its most significant computing power commitment to date to match unprecedented growth."
Client-side data is equally impressive. During the Series G financing in February, there were 500 enterprise customers with annualized spending exceeding $1 million. In less than two months, this number doubled to over 1000. According to Anthropic's previous disclosures, the core driver of this growth is Claude Code, which was launched in May 2025 and reached an annualized revenue of over $2.5 billion by February 2026.
For reference, OpenAI's annualized revenue is estimated by Sacra to be approximately $25 billion (as of February 2026). According to Epoch AI analysis, Anthropic's annualized growth rate since surpassing $1 billion in revenue is about 10x, while OpenAI's is around 3.4x during the same period. At this rate, the revenue crossover point between the two may occur in mid-2026.
Note: The above figures are annualized revenue (run-rate revenue), which is an estimate of recent monthly revenue multiplied by 12, not actual cumulative revenue.
3.5 Gigawatt TPU Agreement: The Latest Piece of Anthropic's Computing Power Strategy
According to Broadcom's SEC filing, the core content of the agreement is as follows: Broadcom will design and supply next-generation TPU chips for Google, with the supply relationship extending to 2031. Anthropic will acquire approximately 3.5 gigawatts of next-generation TPU computing power through Broadcom starting in 2027, as part of its "multi-gigawatt" computing power expansion plan.
Broadcom added a key qualification in the filing: "Anthropic's consumption of expanded computing power depends on its continued commercial success." The three parties are also discussing deployment support with "operational and financial partners."
This is not Anthropic's first large-scale computing power agreement. In October 2025, Anthropic signed a cooperation agreement with Google Cloud, gaining access to up to 1 million TPUs, expected to deliver over 1 gigawatt of computing power in 2026. Broadcom CEO Hock Tan confirmed in a December 2025 earnings call that Anthropic had placed two TPU orders worth $10 billion and $11 billion, respectively. In a March earnings call, Tan further stated that Broadcom expects to generate approximately $21 billion in AI revenue from Anthropic in 2026 and over $42 billion in 2027 (according to Mizuho analyst estimates).
On the AWS side, Project Rainier went live in October 2025, deploying nearly 500,000 Trainium2 chips across multiple data centers in the United States. Amazon has invested a total of $8 billion in Anthropic. Anthropic engineers are directly involved in the development of Trainium's underlying kernels and have provided design input for the next-generation Trainium3 chips.
Thus, Anthropic's computing power sources cover three chip routes (AWS Trainium, Google TPU, and NVIDIA GPU) and three major cloud platforms (AWS, Google Cloud, and Microsoft Azure). Anthropic specifically emphasized in its blog that Claude is "the only frontier AI model available on all three major global cloud platforms."
Diverging Paths from OpenAI's Stargate
Anthropic's computing power model contrasts sharply with OpenAI's.
OpenAI has chosen a capital-intensive route: in January 2025, it co-founded Stargate LLC with SoftBank and Oracle, aiming to invest $500 billion over four years to build 10 gigawatts of AI infrastructure. OpenAI holds operational responsibility and design control, Oracle is responsible for construction, and SoftBank bears financial responsibility. To date, Stargate's planned computing power is close to 7 gigawatts, with cumulative investment commitments exceeding $400 billion.
However, Stargate's progress has been marred by control disputes among partners. According to a Tom's Hardware report in February, disagreements over data center ownership between OpenAI, Oracle, and SoftBank have caused delays in some projects. Additionally, OpenAI's total cloud service procurement commitments exceed $500 billion ($250 billion with Microsoft, approximately $300 billion with Oracle, and about $50 billion with AWS). It is estimated that OpenAI will burn through approximately $17 billion in cash in 2026 and may not achieve positive cash flow until 2030 at the earliest.
Anthropic follows a "no data centers, no chip purchases" asset-light model. Capital expenditures are borne by cloud service providers, while Anthropic, as a customer, locks in capacity and prices through long-term agreements, retaining flexibility to switch between multiple chip routes. The trade-off is that it does not own the infrastructure, potentially leading to higher long-term unit costs. According to reports, Anthropic's gross margin is about 40%, with an estimated loss of approximately $14 billion in 2026.
The superiority of either model remains undetermined. OpenAI is betting on economies of scale and infrastructure autonomy, while Anthropic is betting on supply chain flexibility and capital efficiency. However, one fact is clear: in the AI computing power arms race, long-term computing power agreements are becoming as critical a competitive factor as funding and technology.
