Anthropic’s Best AI Models Struggle to Attract Users, While Cheaper Tools Are Thriving
This Financial Times (FT) report obtained some interesting data from “people familiar with the matter”:
- Anthropic’s “annualized revenue” peaked at $65 billion in July, compared with $47 billion in May. I’ve compiled more historical data here.
- Based on the same model it used to announce second-quarter profitability, Anthropic expects to be profitable in the third quarter as well. “The company also told investors that it has 6,000 customers who spend $100,000 or more each year.”
- As for OpenAI, “annualized revenue has grown 35% so far this quarter and now exceeds $40 billion. GPT 5.6, released in July, boosted the company’s performance after a weak start to the year.”
The article also introduced me to the Ramp AI Index. The index uses billing data from 70,000 businesses that use Ramp credit cards to estimate model adoption.
Here is Ramp’s breakdown of spending on Anthropic’s models in July 2026. Given that Opus 5 was not released until July 24, these results look reasonable and support the idea that Fable’s cost makes it a less popular model:
- Opus 4.8: 28.0%
- Sonnet 4.6: 8.3%
- Fable 5: 8.0%
- Opus 4.6: 6.9%
- Sonnet 5: 3.6%
- Opus 5: 3.5%
- Opus 4.7: 1.7%
- Sonnet 4.5: 1.3%
- Haiku 4.5: 1.0%
- Opus 4.5: 0.7%