Key takeaways
What happened
Anthropic's Fable 5 is considered the most capable AI model on the market. S. companies are barely adopting it. Spending data from financial services provider Ramp shows that companies are barely buying Anthropic's most powerful model through its API. In its first month after launch, Fable 5 accounted for only about six percent of the tokens purchased from Anthropic. 4 percent.
New customers keep signing up with American model providers, but advanced users, whose growing spending OpenAI and Anthropic increasingly depend on, are shifting toward open-source models. Ramp's data shows those models now trail frontier models by only a few months, and as a result, growth at the two leading AI labs is slowing. Despite the skepticism around paying top dollar for premium models, total AI spending keeps rising. S.
Why it matters
6 Sol, captures 25 percent of tokens and 23 percent of spending at OpenAI. 6 Sol generated, despite costing significantly more per token. Ramp notes that the sample for the Fable data comes from the company's proprietary token spend management product and skews slightly toward tech companies. Actual Fable adoption is likely even lower than these estimates, assuming Fable 5 is used mainly for coding.
Ramp economist Ara Kharazian attributes Fable 5's slow uptake to its price. 6 Sol or other Anthropic flagship models. Kharazian sees this as a new ceiling on what companies are willing to spend on AI, arguing that the extra performance simply isn't worth the cost. It's likely more complicated. The performance edge Fable 5 offers may not matter for many use cases, or it's barely measurable in daily work.
This points to a basic problem with calculating AI return on investment. How does a company put a number on the value an AI model delivers, especially when trying to measure the gap between one model generation and the next? It's a complicated and messy equation. But the data doesn't say that a Fable 5 class model represents the upper limit of what companies would pay for AI per se.
Models that are dramatically more capable could also deliver dramatically higher and, more importantly, tangible value. Companies will buy what pays off. But as long as that value stays abstract, their willingness to pay appears to be limited. S. 1 percentage points from the previous month. 23 percentage points, lagging overall AI adoption growth. 94 percentage points to 4 percent.
What to watch
companies spent a median of $7,400 per employee on AI. 95 per employee. Ramp's data suggests companies are spending more on AI, but not without limits. Willingness to pay sharply higher prices for performance gains that are difficult to measure in daily work appears to have plateaued, at least if Fable 5 is any indication.
According to Ramp, these are worrying signs for the AI industry, whose investment thesis depends on fast-growing revenue from increasingly powerful models.