Anthropic's Claude models have long been regarded as a serious challenger to OpenAI's ChatGPT and GPT lineup in the enterprise market. Now, fresh transaction data shows the challenger has taken the lead.

According to Ramp's AI Index for May 2026 — which analyzes corporate card and invoice data from more than 50,000 U.S. companies — Anthropic now controls 34.4 percent of enterprise AI tool spending. OpenAI sits just behind at 32.3 percent. It is the first time Anthropic has led in Ramp's dataset, TechCrunch reports.

What does Ramp actually measure?

It is worth noting what these figures actually say — and what they do not. Ramp defines a "verified enterprise customer" as a business with at least one positive transaction to an AI product or service during a given month. The underlying data consists of anonymized card payments and invoices from more than 70,000 U.S. companies that use Ramp's corporate card and payments platform.

The data captures actual financial commitment to AI tools — not self-reported usage or free trial accounts

The methodology is therefore grounded in real spending, not surveys. The downside is that free AI services and employees using personal accounts for work-related tasks are not captured. The figures should therefore be read as a measure of paid AI adoption in the enterprise market, not total market share.

Anthropic overtakes OpenAI: Now leads in paying enterprise customers - Bilde 1

Strong growth in large contracts

The growth in Anthropic's enterprise base is nonetheless significant from any angle.

300,000+
Total enterprise customers
500+
Customers spending over $1M annually

Two years ago, the company had only "a dozen" customers in the latter category, according to its own figures. Growth in large contracts has reportedly increased nearly sevenfold year over year. Among its customers, the company counts eight of the world's ten largest companies as measured by the Fortune 10 list.

Reuters reported in April 2026 that Anthropic's annualized revenue surpassed $30 billion — a figure that underscores that the growth is not just about breadth of customers, but also depth of contracts.

Claude Code drives growth

A significant driver is Claude Code, Anthropic's coding tool aimed at developers and engineering teams. Weekly active users doubled from the start of the year through February 2026, and the product's run rate surpassed $2.5 billion during the same period.

Companies such as Stripe, Zoom, Honeycomb, and Instacart are actively using Claude Code in their development workflows. Zoom is also an investor in Anthropic through Zoom Ventures and integrates Claude into its own products.

Among companies adopting AI for the first time, Anthropic wins around 70% of head-to-head evaluations against competitors

The broad customer portfolio spans from financial services firms such as Visa, Stripe, and LSEG, to design tools like Canva and fitness chain Equinox. Equinox uses Claude, among other things, for personalized fitness and nutrition recommendations in its mobile app.

Defense and intelligence contracts

Anthropics ambitions also extend into the U.S. defense establishment. In partnership with Palantir and Amazon Web Services, the company offers the Claude model to U.S. intelligence and defense agencies. A dedicated version called "Claude Gov" is in active use at several national security agencies. The U.S. Department of Defense signed a contract worth $200 million in July 2025 for AI in the military, with Anthropic among the vendors.

Anthropics lead in Ramp's enterprise measurement is a concrete data point, but not necessarily a definitive answer to who is winning the enterprise AI race. OpenAI still commands a massive consumer base and strong partnerships. What is clear is that the competition between the two companies is now closer than ever — and that businesses are voting with their credit cards.