According to Slashgear , generating just one AI image consumes about 0.011 kilowatt-hours (kWh) of electricity. That might sound small, but it’s roughly the same amount of energy your refrigerator uses in four minutes.

To put it in perspective, a standard fridge running continuously for a full day consumes about 4 kWh, according to EnergySage .

That means every AI-generated picture carries a tangible, if tiny, environmental footprint, one that multiplies quickly when millions of images are created daily across platforms.

While a fridge hums quietly in the background, the servers powering AI models run at massive data centres, drawing electricity from grids that often rely on fossil fuels.

Slashgear notes that a single AI image may require dozens of processing steps from high-performance GPUs, each contributing to that energy cost.

Multiply this by the growing number of users experimenting with tools like Midjourney, DALL·E, or Stable Diffusion, and the numbers add up quickly.

Globally, the conversation around AI’s energy footprint is intensifying. The UN Environment Programme recently highlighted that as AI adoption accelerates, so does its electricity demand, raising concerns about sustainability and carbon emissions.

With major tech companies investing billions into generative AI, the race isn’t just about smarter models anymore; it’s also about greener ones.