How Much Water Does a ChatGPT Query Really Use?
With AI use skyrocketing, many are starting to ask a critical question: how much water does a ChatGPT query use? In a recent blog post, OpenAI CEO Sam Altman claimed that a single ChatGPT request consumes about one-fifteenth of a teaspoon of water. That’s approximately 0.000085 gallons per query. While the number may seem tiny, it adds up quickly when you consider the billions of queries processed daily. Altman’s statement touches on growing environmental concerns around AI energy and water usage—questions that are becoming more relevant than ever in 2025.
ChatGPT Water Usage Explained by Sam Altman
In his post, Altman compared the energy and water consumption of a ChatGPT request to familiar household items. He noted that the average ChatGPT query uses around 0.34 watt-hours of energy—similar to what an oven uses in just over a second. The water usage, while minimal on a per-query basis, becomes notable when scaled globally. Though Altman didn’t cite a specific source for his water estimate, his point was clear: as AI scales, its environmental footprint can’t be ignored.
Why AI Water Consumption Is Sparking Debate
The water needed to cool massive data centers is now under the microscope. Reports from 2024 and 2025 warn that AI could soon consume more electricity than Bitcoin mining, with a similar rise in water use. Investigations like one from The Washington Post revealed that AI-generated content, even short outputs like a 100-word email, can use water equivalent to an entire bottle. Factors like datacenter location and cooling tech significantly impact these figures. Altman’s statement might be an attempt to get ahead of public concern and steer the narrative.
Is AI Sustainability the Next Big Challenge?
As users, developers, and policymakers push for sustainable AI, transparency around water and energy use is crucial. Sam Altman’s teaspoon comparison might sound dismissive, but it also brings attention to a vital issue: the hidden environmental costs of AI. With more people using tools like ChatGPT daily, understanding and optimizing resource use is no longer optional—it’s essential for responsible innovation in tech.