- Published on
OpenAI GPT-5.6 Sol Revenue Overtakes Anthropic Fable 5
- Authors
- Name
- DeepHorus
According to the latest AI Index data released by fintech company Ramp, in July this year, U.S. companies' spending on OpenAI's GPT-5.6 Sol model was approximately 1.33 times that of Anthropic's Fable 5 model, although Anthropic still leads in enterprise adoption metrics. This data offers a glimmer of hope for OpenAI, while also raising concerns about a slowdown in corporate AI spending growth.
Ramp analyzes the competitive landscape of AI service providers by tracking billing and spending data from over 70,000 U.S. corporate customers. Although its customer base skews toward small and medium-sized enterprises, the index is still regarded as an important reference in the industry. The latest data shows that Anthropic's top model has become less popular among enterprises compared to OpenAI's corresponding product.
Anthropic's Fable 5 model had a rocky rollout. The model was temporarily taken down due to government orders, only returning to service on July 1, and its price was adjusted multiple times throughout the month. Meanwhile, OpenAI's cheaper GPT-5.6 Sol is forming a competitive threat, which may be one of the reasons for its relative lag in enterprise spending.
On the eve of its IPO, Anthropic is betting that users will pay higher prices for the most advanced AI technology rather than choosing cheaper alternatives. However, the latest data suggests that enterprise customers may be controlling their AI spending, posing a potential challenge to Anthropic's business prospects.
Despite lagging in top-model revenue, Anthropic still leads in enterprise adoption. In July, the share of U.S. companies willing to pay for Anthropic's AI products reached 43.5%, higher than OpenAI's 39.7%. xAI's enterprise adoption rose to 4%, while Google's fell from 6.4% to 6.2%.
Ramp economist Ara Kharazian noted that Google's actual data may be underestimated because some companies use AI products through Google Workspace integration. He also said that the popularity of Chinese open-source AI models has increased, but not yet to the point of making OpenAI and Anthropic 'panic.'
As enterprises gradually adopt AI products, the growth rate of AI spending has accelerated significantly. In July, among the top 1% of companies with the highest AI spending, the median spending per employee on AI tools reached $7,400, up from less than $5,000 in June. The overall median AI spending per employee across companies also rose from about $11 to $12.
AI agents have become an important driver of revenue growth in the AI industry this year, and are also a key R&D focus for both OpenAI and Anthropic. Kharazian said that a large portion of the AI spending tracked by Ramp comes from coding agents, which can automatically trigger more tasks, causing usage costs to 'snowball.' 'This is why spending grows so quickly: it has a scalability that traditional software lacks,' he said.
