Anthropic’s Fable 5 Has 11% of Enterprise Spend as Cheaper Opus 5 Pulls Ahead

Anthropic’s most advanced model is not leading its enterprise spending data. Two months after launch, Fable 5 accounted for about 11% of enterprise spending on Anthropic models, while the cheaper Opus 5 had surpassed it after its late-July debut. The gap arrives as Anthropic is reportedly preparing for a $2 trillion-or-more IPO.

A premium launch meets a different buying test

The signal comes from Ramp payment data cited by the Financial Times. It breaks with what CIO described as a prior pattern in which enterprise customers moved quickly to Anthropic’s most powerful model. Analysts and investors cited in that account attributed Fable 5’s slower start mainly to its higher price and to cheaper models being sufficient for many tasks.

Opus 5 is the immediate counterpoint. The less expensive model surpassed Fable 5 in enterprise spending after launching in late July, according to the same payment-data account. That comparison is about spending rather than a full measure of model use or capability, but it places price alongside raw performance in the enterprise buying decision.

Growth remains the other side of the ledger

The payment pattern sits beside unusually fast reported revenue growth. Anthropic’s annualized revenue run rate reportedly rose from roughly $9 billion at the end of 2025 to $47 billion in May and more than $65 billion by the end of July. Anthropic also reportedly told investors it had 6,000 customers spending at least $100,000 annually.

The company reportedly recorded its first adjusted operating profit in the second quarter and expected to remain profitable in the third. Reuters, as cited in the 24/7 Wall St. account, said Anthropic was projecting roughly $190 billion to $200 billion in 2028 revenue. At a $2 trillion valuation and $200 billion in revenue, that would equate to roughly 10 times sales.

The next move is a pricing-and-demand test

The early Fable 5 result does not establish that Anthropic’s overall growth has slowed; its reported run rate moved sharply higher between May and July. It does, however, put a narrower issue in focus: whether each new frontier model can translate improved capability into higher customer spending, rather than shift work toward a cheaper tier.

Competitive pricing raises the stakes. The 24/7 Wall St. account said OpenAI’s annualized revenue had climbed above $40 billion and that its GPT 5.6 models were priced below Fable 5. Anthropic declined to comment on the payment data, leaving the next enterprise-spending readings as a consequential test of whether Fable 5’s start is temporary or a more durable preference for lower-cost models.