Soaring bills reshape how businesses choose AI models - iTnews

An article from iTnews reports that while the price of individual tokens has been decreasing, the overall cost of completing AI tasks is increasing. This is due to a shift by AI providers toward usage-based pricing models, where charges accumulate based on the number of tokens processed per task. As a result, businesses are reconsidering their model choices, with many opting for cheaper, less capable models to manage expenses.

The trend reflects a broader change in the AI market dynamics. In response, enterprises are evaluating total cost of ownership more carefully, sometimes switching to open-source models or smaller proprietary alternatives that offer sufficient performance at lower cost.

The shift may accelerate commoditization of AI inference services, benefiting model optimizers and cost-conscious adopters. It also pressures premium models to demonstrate clearer value differentiation. The article concludes by noting that the market for AI models is becoming more segmented, with cost becoming a primary decision factor alongside capability.