For years, flat-rate AI subscriptions heavily subsidized power users, masking the true cost of token consumption. As providers shift to usage-based billing, this subsidy is ending. Enterprises now face "sticker shock" and must justify AI spend with clear ROI, moving from rampant experimentation to cost-conscious implementation.

For the first time in years, the perceived leap in LLM capabilities has slowed. While models have improved, the cost increase (from $20 to $200/month for top-tier access) is not matched by a proportional increase in practical utility, suggesting a potential plateau or diminishing returns.

The AI market has cleared its first ROI hurdle: model revenue has justified massive infrastructure investment. Now it faces a second, harder test. Enterprises spending billions on AI tokens must demonstrate tangible financial benefits, like higher margins or revenue, to sustain the flywheel.

The most heated topic among Fortune 500 CIOs is no longer which AI model is most powerful, but how to manage unpredictable and soaring token costs. Companies are struggling to find the right strategies—from workload prioritization to user-based access tiers—to create a predictable cost model in a rapidly evolving tech landscape.

An anecdote about an engineer spending $100M in a month on AI tokens reveals a core enterprise issue. For Lenovo's CFO, the problem isn't the amount but its lack of planning and clear ROI. This signals a shift from predictable software subscriptions to volatile, usage-based AI compute costs.

While early generative AI costs were negligible, the shift to complex, multi-step agentic workflows is causing a massive spike in token usage. This has elevated cost optimization and ROI from a minor concern to a C-suite priority for the first time.

The primary short-term risk for the AI sector isn't capital expenditure but the high cost of token generation. For AI applications to become ubiquitous, the unit economics must improve. If running a single query remains prohibitively expensive for businesses, widespread, sustainable adoption will be impossible, threatening the entire investment thesis.

While the cost to achieve a fixed capability level (e.g., GPT-4 at launch) has dropped over 100x, overall enterprise spending is increasing. This paradox is explained by powerful multipliers: demand for frontier models, longer reasoning chains, and multi-step agentic workflows that consume exponentially more tokens.

The current affordability of AI tokens is not sustainable; it's propped up by venture capital funding AI companies operating at a loss. Businesses should treat this as a temporary window for aggressive learning and experimentation before prices inevitably rise to reflect true operational costs.