Anthropic's breakout month
Why it may be a while before we declare a winner in the AI race
Dear Colleagues: Today’s letter includes my monthly update of Ramp AI Index, our flagship research using spend data from Ramp to track how American businesses are using AI. As a subscriber, you will get my analysis before anyone else. I post on Substack first on the second Wednesday of every month at 10 a.m.
Extras from Ramp data:
The Wall Street Journal covered our findings that Anthropic commands 80% of API spend vs. OpenAI on Ramp. (WSJ / X)
11% of startups applying to Ramp are AI-positioned (using a .ai domain or mentioning AI in their company description). (LinkedIn)
6 of the trending companies on our Top Software list sell AI infrastructure (X). My dear colleague Ryan Stevens just published a deep dive into this rapidly growing, surprisingly narrow market.
Ramp AI Index shows overall business AI adoption rose to 46.8% of businesses in January, a new high, with Anthropic surging and now used by 19.5% of businesses.
Anthropic adoption grew from 16.7% to 19.5%, one of its largest monthly gains since we started tracking. One in five businesses on Ramp now pays for Anthropic (a year ago, it was one in 25). Meanwhile, OpenAI adoption slipped from 36.8% to 35.9%, giving back some of last month’s gains. Google held steady at 4.5%.
The natural question is whether Anthropic is gaining at OpenAI’s expense. I’ve been hesitant to call a winner in the AI race — the market is too nascent and dynamic. The asks to do so will increase with this post, but I see three scenarios worthy of investigation.
One of the following is true:
OpenAI’s adoption rate is falling (while Anthropic’s is growing) because businesses are cancelling their OpenAI subscriptions and contracts, and switching to Anthropic.
There is no switching behavior, and Anthropic’s growth is driven by existing OpenAI customers adding Anthropic’s services alongside OpenAI
New AI adopters, perhaps newly established startups or established businesses that have never bought AI before, are picking Anthropic over OpenAI from the get-go.
Are businesses replacing OpenAI with Anthropic, or are they buying both?
We can rule out the first scenario quickly. OpenAI and Anthropic have nearly identical churn rates (around 4% of businesses cancel their subscriptions or contracts a month). If businesses were switching from OpenAI to Anthropic, you’d see that in OpenAI’s churn. You don’t.
The data points to Scenario 2 because of one key metric: overlapping customer bases. Ramp data shows about 79% of Anthropic’s customers also pay for OpenAI. Anthropic is adding new users faster, but those users aren’t first-time AI buyers. They’re existing OpenAI customers adding a second provider.
So who is winning the AI race?
The findings suggest the race isn’t zero-sum — at least not yet. The market is young enough that businesses are buying from more than one model company. Different tools for different teams, different tasks. Engineers prefer one model, the sales team uses another, and the company pays for both.
Another read is that businesses are experimenting with Anthropic as a second vendor and will eventually pick a winner. In this scenario, the overlap will shrink over time as firms consolidate on one vendor.
Today, 16% of businesses pay for both OpenAI and Anthropic. A year ago it was 8%. This overlap metric is often lost in simple market share analyses, but it may be the most important one to watch.
Something from my camera roll
STAND AND DELIVER



interesting and all but what is ramps payment volume vs some of the large bank or startup competitors? Like how much of the market does this represent
great analysis