Fable 5's weak sales signal ceiling on what companies will pay for AI

Anthropic's Fable 5 launched to considerable fanfare as the most capable AI model on the market, but sales data from financial services provider Ramp tell a starkly different story. In its first month after launch, Fable 5 accounted for only about 6 percent of the tokens purchased from Anthropic through its API, and 11.4 percent of total spending on Anthropic models. By contrast, OpenAI's flagship model, GPT-5.6 Sol, captures 25 percent of tokens and 23 percent of spending at OpenAI—and generates about 75 percent more revenue than Fable 5, even though Fable 5 costs significantly more per token. Fable 5 is priced at about $10 per million input tokens and $50 per million output tokens, roughly double the cost of GPT-5.6 Sol or other Anthropic flagship models.

Ramp economist Ara Kharazian attributes this weak adoption to price, arguing that companies have discovered a ceiling on what they are willing to spend for AI performance gains that are difficult to quantify in everyday work. The company notes that calculating AI return on investment remains murky—how do you put a dollar value on the performance difference between one model generation and the next?—and that many use cases may not benefit enough from Fable 5's marginal edge to justify the premium. The sample underlying Ramp's data skews slightly toward tech companies and comes from its proprietary token spend management product, suggesting actual Fable 5 adoption is likely even lower.

Corporate AI spending overall continues to rise: in July, the top 1 percent of U.S. companies spent a median of $7,400 per employee on AI, the top 10 percent spent $650, and the median company spent $11.95 per employee. Yet this growth masks troubling trends for the major AI labs. Anthropic reached 43.5 percent adoption among U.S. companies in July (up 1.1 percentage points month-on-month) and has passed OpenAI, but OpenAI's growth has slowed to just 0.23 percentage points. More significantly, advanced users—whose growing spending OpenAI and Anthropic increasingly depend on—are shifting toward open-source models, which now trail frontier models by only a few months. xAI posted faster growth than both, rising 0.94 percentage points to 4 percent, though it remains a niche provider. Ramp characterizes these trends as worrying signs for an AI industry whose investment thesis depends on fast-growing revenue from increasingly powerful models, and suggests that the willingness to pay sharply higher prices for performance gains that are hard to measure has, at least for now, plateaued.