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

← All comparisons

ChatGPT vs a Google search: energy and water compared

A classic Google search uses about 0.3 Wh; a ChatGPT prompt at our Mid scenario about 1.4 Wh: roughly 4-5× the energy, and a few times the water. The gap was 10× in 2023 headlines; efficiency gains and Google's own AI answers have blurred it since.

One AI prompt (Mid)

7 mL

one Google search

1.5 mL

Prompt: one typical ChatGPT exchange at the Mid scenario (cooling plus the water behind the electricity). Item: its full water footprint, as published by the source.

One AI prompt uses as much water as

≈ 4.7 ×

one Google search

1 dot = 1 × one Google search

The '10× a Google search' line traces to a 2009-era Google figure (0.3 Wh per search) against early ChatGPT estimates near 3 Wh. Both ends have moved: current per-query estimates for standard chatbot use cluster well under 1 Wh (Google's median Gemini prompt: 0.24 Wh; Altman's ChatGPT figure: 0.34 Wh), while search itself now often runs an AI answer on top.

Our frontier-class Mid estimate for a ~700-token ChatGPT exchange is about 1.4 Wh; call it four to five searches. Reasoning modes change the picture by an order of magnitude, which is why the calculator has a separate class for them.

The deeper point: the marginal cost of asking either tool one more question is tiny; the systemic cost is the build-out both are driving. That story is in TWh and watersheds, not per-query milliliters.

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. 1Jegham, 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
  2. 2Altman, S. (2025). The Gentle Singularity. blog.samaltman.com: 0.34 Wh and 0.000085 gal of water per average ChatGPT query. (accessed 2026-06) official disclosure
  3. 3Siddik, M. A. B., Shehabi, A., & Marston, L. (2021). The environmental footprint of data centers in the United States. Environmental Research Letters 16(6): watershed-scale direct + indirect water footprint. (accessed 2026-06) peer-reviewed
  4. 4Google (2025). Measuring the environmental impact of AI inference. Google Cloud technical disclosure: 0.24 Wh / 0.26 mL / 0.03 gCO₂e per median Gemini Apps prompt. (accessed 2026-06) official disclosure
  5. 5You, J. (Epoch AI) (2025). How much energy does ChatGPT use?. Epoch AI Gradient Updates: GPT-4o per-query energy estimate ~0.3 Wh. (accessed 2026-06) industry report