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.

As corporate spending on AI rises rapidly, the big model-centric structure of the AI industry is showing signs of change. Developers have kept inference costs relatively low using invested funds, allowing companies to use advanced models below cost. With AI adoption spreading, more companies say costs are becoming a burden. Coinbase CEO Brian Armstrong forecast most work will shift to much cheaper models, a shift that could affect profitability at major AI firms.