Skip to content
AIWaterUse.org
ES
Menu
↑ ↓ to move · Enter to open · Esc to close

xAI

How much water does xAI's Grok use?

xAI built the world's largest single AI training cluster in Memphis in months. Local reporting puts its water draw around a million gallons a day.

~1M gal/day

estimated water draw, Colossus cluster, Memphis (local reporting)

xAI is the speed-run version of the data center story: the Colossus cluster in Memphis went from empty industrial shell to hundreds of thousands of GPUs in months, fast enough that power came partly from on-site gas turbines while grid connections caught up.

xAI publishes no environmental reporting, so the public record is local: utility filings and Memphis journalism put the site's water demand on the order of a million gallons per day, in a city whose drinking water comes from the Memphis Sand aquifer. A planned greywater recycling facility is meant to reduce the potable share; whether and when it does is exactly the kind of fact local reporting tracks and ESG reports don't.

Grok itself has no published per-query figures; we class it as frontier. The honest summary: per prompt, Grok is probably ordinary. As a neighbor, Colossus is the most-watched data center in America.

Your number

What does your AI actually use?

Pick your model, set your usage, get the number, with sources.

Try the calculator

Sources

Cited on this page, in order of appearance.

  1. 1Local utility filings and news reporting (2025). Aggregated local reporting on individual data center campuses. Utility filings and local journalism on per-site water demand (Memphis, West Des Moines, The Dalles, Bluffdale, metro Atlanta), used where ESG reports do not break sites out. (accessed 2026-09) news
  2. 2Shehabi, A., Smith, S. J., Hubbard, A., et al. (2024). 2024 United States Data Center Energy Usage Report. Lawrence Berkeley National Laboratory, LBNL-2001637: 176 TWh electricity and 17B gal direct water in 2023, with 2028 projections. (accessed 2026-06) government
  3. 3Jegham, N., Abdelatti, M., Elmoubarki, L., & Hendawi, A. (2025). How Hungry is AI? Benchmarking Energy, Water, and Carbon Footprint of LLM Inference. arXiv preprint arXiv:2505.09598 (v6, Nov 2025). (accessed 2026-06) preprint