More than half of Anthropic’s annual enterprise customers now use both the Claude platform and at least one first-party application, a figure that has risen from roughly 30% at the end of last year. The company’s S-1 filing, submitted to the Securities and Exchange Commission on Monday, states that this adoption rate is proof of its standing at the frontier of artificial intelligence.
As of September 30, the firm recorded over 11,900 customers spending at least $100,000 annually. This group includes more than 1,900 clients spending over $1 million and approximately 250 spending $10 million or more each year.
Expanding beyond the API
Anthropic credits this momentum to customers realising the value of deploying Claude across their whole organisations. This shift drove net revenue retention to about 900% by the end of the third quarter.
The filing rejects the idea of Claude as a static tool for a single task. Instead, it describes the offering as frontier AI with rapidly advancing capabilities spanning the models, the developer platform, and the applications users interact with directly. The document argues that an advance in any one of these areas improves the others, accelerating the delivery of more capable intelligence. This integrated approach is cited as the reason more clients are adopting both the platform and first-party apps.
The strategy mirrors established tech giants. Uber started by selling rides before expanding into food delivery and freight. Amazon used cloud computing to attract users before adding storage and databases to make the ecosystem harder to leave. Similarly, a business that begins with the API can deploy Claude Code to engineers and then Cowork to the wider workforce.
Customers using these applications become significantly more valuable. In the third quarter, the top 5% of enterprise clients who used Claude Code spent about 20 times more per month than the average user who did not use the coding tool. This data shows the potential for higher per-user spend and explains why Anthropic wants its applications inside accounts that start with the API.
To encourage this organic spread, Anthropic invests in Applied AI teams. One internal team works with startups and enterprises to place Claude inside existing technology stacks. The company also launched Ode with Anthropic, a venture with Blackstone and other private equity firms, to embed forward-deployed engineers into businesses to build bespoke AI tools.
Distribution partnerships with AWS, Google Cloud, and Microsoft Azure also play a role. Under a joint go-to-market approach, Anthropic maintains the direct contractual relationship with customers and pays each cloud provider a fee. The company says this allows it to avoid channel conflict and set its own pricing.
This stickiness strategy may become essential as the firm seeks to constantly reinvent its business. Anthropic plans to release new classes of products that dwarf the economic value of its existing ones. The company intends to devote less effort to entrenching current products and more to defining new ones.
Risks and headwinds
Despite revenue growth of $16.3 billion for the first half of 2026, up from $1.2 billion in the same period last year, the filing flags significant risks. The cloud partnerships that help sell Claude are increasingly seeing Anthropic as a competitor. The document states this could lead to difficulties securing the compute needed from them. Amazon, Google, and Microsoft may prioritise their own requirements or rivals they have invested in.
Customer concentration remains another risk. The risk factors section notes that any adverse change in demand from the largest customers could adversely affect the business and prospects. A significant reduction in spending by, or the loss of, one or more of the largest customers could hurt financial condition and results of operations unless offset by new relationships or increased business from other existing clients.
Anthropic did not immediately respond to a request for comment from TechCrunch.
What it means
For businesses using the technology, the shift means moving from buying a single model to integrating a suite of tools. Users can expect deeper integration between coding assistants, workplace collaboration apps, and the core AI platform, creating a system that is harder to replace. However, reliance on a few large clients and the threat of cloud providers restricting access to computing power remain serious concerns for the company’s future stability.
