Fable 5: Anthropic's AI Struggles to Attract American Businesses

Fable 5: Anthropic's AI Struggles to Attract American Businesses

The slow adoption of Fable 5 suggests that companies' willingness to pay for cutting-edge AI has reached a ceiling.

Anthropic's Fable 5 is considered the most advanced AI model on the market. However, new sales data shows that American companies are barely adopting it.

Spending data provided by financial services provider Ramp indicates that companies are purchasing Anthropic's most powerful model through its API at a minimal rate. In its first month after launch, Fable 5 accounted for only about six percent of the tokens purchased from Anthropic. Measured against total spending on Anthropic models, this figure rose to 11.4 percent.

OpenAI's flagship model, GPT-5.6 Sol, captures 25 percent of the tokens and 23 percent of the spending at OpenAI. Overall, according to Ramp, Fable 5 generated only 75 percent of the revenue related to models that GPT-5.6 Sol produced, despite a significantly higher cost per token.

Fable 5 represents just a fraction of corporate spending on Anthropic models.

Ramp notes that the sample data on Fable comes from the company's token spending management product and slightly favors tech companies. The actual adoption of Fable is likely even lower than these estimates, assuming that Fable 5 is primarily used for coding.

Price May Have Hit a Ceiling for Corporate AI Spending

Ramp economist Ara Kharazian attributes the slow adoption of Fable 5 to its price. The model costs about $10 per million input tokens and $50 per million output tokens, making it roughly twice as expensive as GPT-5.6 Sol or other flagship models from Anthropic. Kharazian views this as a new ceiling on what companies are willing to spend on AI, arguing that the additional performance simply isn't worth the cost.

The situation is likely more complex. The performance advantage that Fable 5 offers may not be relevant for many use cases, or it is barely measurable in day-to-day work. This raises a fundamental issue regarding the calculation of AI return on investment. How can a company quantify the value that an AI model brings, especially when it comes to measuring the gap between one model generation and the next? It's a complicated and delicate equation.

However, the data does not suggest that a model of Fable 5's class represents the upper limit of what companies would be willing to pay for AI per se. Much higher-performing models could also offer tangible and, more importantly, measurable value. Companies will buy what is profitable. But as long as that value remains abstract, their willingness to pay seems limited.

Growth of OpenAI and Anthropic Slows

According to Ramp, 43.5 percent of American companies paid for subscriptions or tokens from Anthropic in July, up 1.1 percentage points from the previous month. OpenAI reached 39.7 percent but only increased by 0.23 percentage points, lagging behind the overall growth in AI adoption. xAI recorded its strongest growth since July 2025, increasing by 0.94 percentage points to reach 4 percent.

New customers continue to sign up with American model providers, but advanced users, whose increasing spending increasingly relies on OpenAI and Anthropic, are turning to open-source models. Ramp's data shows that these models are now only a few months behind the leading models, and consequently, the growth of the two major AI labs is slowing.

Anthropic has surpassed OpenAI in adoption among American companies and is widening its lead. xAI is growing but remains a niche provider.

Despite skepticism around paying high prices for premium models, total spending on AI continues to rise. In July, 1 percent of American companies spent an average of $7,400 per employee on AI. For the top 10 percent, this figure was $650, while the median company spent $11.95 per employee.

Ramp's data suggests that companies are spending more on AI, but not without limits. The willingness to pay significantly higher prices for performance gains that are difficult to measure in day-to-day work seems to have plateaued, at least if Fable 5 is any indicator. According to Ramp, these are concerning signs for the AI industry, whose investment thesis relies on rapid revenue growth from increasingly powerful models.