AI Power Consumption and Electricity Costs

AI training and inference can increase electricity usage compared to traditional computing. Training large models requires significant compute power over extended periods; inference—running models to answer queries—also consumes electricity at scale. This has drawn attention to how AI infrastructure affects electricity demand and costs.

Electricity price differences by state matter economically. Operators building or expanding AI infrastructure may consider electricity costs when choosing locations. Higher electricity prices can increase operating costs; lower prices can reduce them. This affects infrastructure planning, investment decisions, and the affordability of power for households and businesses in regions with growing electricity demand.

Rising demand can influence cost discussions, infrastructure planning, and affordability. This site focuses on electricity prices—our state-by-state data helps explain why prices vary and how they matter for different stakeholders.