The current price represents a significant decline from the high of $2.07 recorded in late May, meaning the cost of running inquiries on popular platforms like ChatGPT, Claude, and Gemini has more than halved in just a few months.
This price collapse is driven by a surge in competition and improved efficiency in AI infrastructure.
Silicon Data and Syz Group attribute the downward pressure to the rise of open-source Chinese models, such as Moonshot’s Kimi K3, which offer cheaper alternatives to established "frontier labs" or leading AI research firms.
To stay competitive, major players like OpenAI have slashed prices for their latest models, while others have adopted dynamic pricing that adjusts based on real-time demand.
While these lower costs benefit developers and businesses building AI applications, they reduce the pricing power and revenue potential for the companies creating the underlying models.
The trend toward cheaper AI could reshape the business landscape for industry leaders like OpenAI and Anthropic, both of which filed for initial public offerings (IPOs) earlier this summer.
With token prices deflating, these companies may face increased pressure to prove their profitability to potential investors.
Furthermore, the drop suggests that the massive capital expenditure—the billions of dollars spent by companies like Microsoft and Nvidia on data centers and AI chips—has created an ample supply of AI capability that may already meet the needs of most users.