TECH Signal 349
Agentic AI costs to rise fivefold by 2028 as complex workflows consume more tokens
↗
Gartner warns that agentic AI inference costs will increase more than fivefold by 2028 because complex workflows consume far more tokens, offsetting savings from cheaper models.
For engineers building AI agents, the cost per task will rise sharply even as model prices fall, because agentic reasoning requires many more inference calls. This shifts the economics toward careful model selection and optimization of routing and orchestration to avoid runaway token consumption. The move to usage-based billing makes these costs harder to predict.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
Gartner forecasts agentic AI costs to increase more than fivefold by the end of 2028.
Cheaper tokens are offset by complex workflows that consume many more tokens, driving up overall inference costs.
Gartner recommends optimizing inference, routing, and orchestration, and assigning each task to the most cost-efficient model capable of handling it.
THE CLUSTER