Enterprise AI Spend Under Scrutiny Despite Falling Costs
The productivity gains and return on investment from AI applications are becoming key factors for businesses, even as token prices decline rapidly.
Despite the rapid fall in artificial intelligence (AI) usage costs, companies are beginning to scrutinize their investments more closely, demanding stronger proof of returns.
The market is seeing a consistent decrease in token prices, with one million input tokens for DeepSeek V4 Flash costing $0.14, OpenAI's GPT-5.6 Luna at $0.20, and Meta's Muse Spark at $1.25. While falling prices might suggest increased AI adoption, enterprises are now focusing on the profitability and efficiency of their AI investments.
Tech leaders are shifting their focus from token prices to productivity increases, revenue growth, and business outcomes. They are now budgeting for AI features, closely monitoring token consumption by teams, and deciding which tasks require expensive frontier models versus cheaper alternatives.
The bar for proving value is rising as AI costs decrease. Companies are examining AI's unit economics before production, capping usage with monthly budgets, and logging every transaction against original estimates. AI is increasingly treated as a budget request rather than an open-ended experiment.
According to tech leaders, model efficiency and falling token prices have already reduced software development costs by up to 30%. The larger shift, however, involves organizations rethinking how AI projects are approved, measured, and tied to business results, moving beyond simply finding cheaper ways to do the same work.