As AI costs rise for large companies and startups alike, price is emerging as a major variable in the AI market alongside performance. Demand for cheaper models is increasing, putting pressure on leading firms such as OpenAI and Anthropic, which are seeking IPOs as early as this year, to cut prices. Companies are shifting workloads to lower-cost or open-source models, while interest grows in tools that let firms switch between external models, in-house systems and open source depending on needs.

U.S. and Chinese AI companies are rolling out new models, and observers say pricing could become a key battleground if performance differences narrow. The Financial Times cited low-cost plans from Chinese startup ZhipuAI and said prices may be set by the cheapest model that is still usable. The report said China’s leading LLM firms are performing at similar levels to U.S. rivals on benchmarks, even as U.S. firms retain higher valuations and advantages in revenue base and distribution.