AI Water Footprint Calculator (ChatGPT)

Calculate the water used to cool data centers when you query ChatGPT, GPT-4 or other LLMs.

Nature 4 model presets Comparisons
Rate this calculator

AI Water Footprint Calculator

Berkeley / UC Riverside cooling-water estimates

Instructions — AI Water Footprint Calculator (ChatGPT)

1

Pick a model

GPT-4 / Claude (large LLM) uses ≈5 mL per query. GPT-3.5 mid-size models use ≈2 mL. Gemini and smaller models use ≈0.3 mL. Traditional Google search uses about 0.2 mL.

2

Choose cooling method

Most data centers use water cooling (full water cost). Air-cooled data centers use about 30% of the water on-site. Immersion cooling (newest) cuts water use to about 5%.

3

Enter query volume

Type how many queries (or use the quick picks). The result shows total liters consumed, plus comparisons to water bottles, showers, toilet flushes, and coffee cups.

Why so much water? Server racks generate heat. Cooling towers evaporate water to dump that heat — about 1.9 L per kWh of compute on average (Green Grid, 2024).
It is shifting: Newer models are more efficient. UC Riverside (2023) estimated 500 mL per query; OpenAI in 2025 claims 0.3 mL. We use 5 mL as a mid-range estimate.

Formulas

Water consumption is the product of compute energy and the data center's Water Usage Effectiveness (WUE). Total water per query combines on-site cooling water and off-site water used to generate the electricity.

Total Water per Query
$$ W_{query} = w_{query} \times C_{cooling} $$
w is the water per query at standard water cooling (mL). C is a multiplier for the cooling method: 1.0 water, 0.3 air, 0.05 immersion.
Cumulative Water
$$ W_{total} = N \times W_{query} $$
N is the number of queries. Multiply per-query water by the count to get total liters.
Water Usage Effectiveness
$$ WUE = \frac{W_{total}}{E_{total}} \;\text{[L/kWh]} $$
Industry average ≈ 1.9 L/kWh (Green Grid 2024). Google's 2023 fleet average WUE was 0.95 L/kWh; Microsoft was 0.49 L/kWh.
Energy per Query (LLM)
$$ E_{query} = \frac{T \times P_{token}}{1000}\;\text{[Wh]} $$
T is the number of tokens (input + output). P_token is the energy per 100 tokens, around 0.09 Wh for GPT-4-class models.
Bottles Equivalent
$$ N_{bottles} = \frac{W_{total}}{0.5\,\text{L}} = 2 \times W_{total} $$
Standard 500-mL plastic water bottles. 100 queries at 5 mL ≈ 0.5 L = 1 bottle.
Showers Equivalent
$$ N_{showers} = \frac{W_{total}}{100\,\text{L}} $$
10-minute shower at 10 L/min = 100 L. 10,000 queries ≈ 0.5 showers.

Reference

Quick Reference — Water per Query (mL)
ServicePer querySource
UC Riverside GPT-3 (2023)500 mLOriginal water-footprint paper
GPT-4 / Claude (large)5 mLMid-range estimate
GPT-3.52 mLMid-range estimate
OpenAI 2025 claim0.3 mLSam Altman public statement
Gemini (Google 2025)0.26 mLGoogle data center average
Google Search~0.2 mLIndustry estimate

Scale of AI water consumption

Big AI water moments
EventWater
Training GPT-3 (2020)~700,000 L
Training GPT-4 (2023)~1,500,000 L (est)
US DC water (2014)21.2 bn L/yr
US DC water (2023)66 bn L/yr
Projected AI (2027)4.2–6.6 bn m³/yr
Olympic pool2,500,000 L
Cooling methods
MethodWUE (L/kWh)
Air cooling0.3–0.5
Water cooling2.0–5.0
Immersion cooling~0.1
Hybrid (modern)1.0–2.0
Free cooling (cold climate)0.3
Industry avg1.9

About 80% of AI's total water footprint comes from electricity generation off-site (power plants need water for cooling), not from data-center cooling itself.

Article — AI Water Footprint Calculator (ChatGPT)

AI Water Footprint Calculator

An AI water footprint counts the fresh water used to run an AI model: roughly 0.3 to 5 mL per ChatGPT-class query at a typical data center, depending on model size and cooling method. UC Riverside's 2023 paper estimated 519 mL per 100-token GPT-3 query. OpenAI's 2025 figure is 0.3 mL. The truth sits in between — between 0.3 and 25 mL per query, with most modern estimates at a few mL.

About 80% of that water is consumed off-site, at the power plants that generate electricity for the data center. The remaining 20% is direct cooling water that evaporates from the data center's cooling towers. Training a large model like GPT-3 used about 700,000 liters in 2020 — the annual water consumption of 28 average Americans.

What is the AI water footprint?

The AI water footprint is the total fresh water consumed by AI systems. It splits into two parts: the water used during model training (one-time, large) and the water used during inference (per query, ongoing). Both happen inside data centers that need cooling.

Cooling towers evaporate water to dump heat. Each evaporated liter takes about 2.3 megajoules of heat with it. A typical data center has a Water Usage Effectiveness (WUE) of 1.9 L per kWh of compute — meaning every kilowatt-hour of electricity translates to 1.9 liters of water gone. The 2023 industry average is from the Green Grid consortium.

Did you know

The first major study to quantify AI water footprint was UC Riverside's 2023 paper "Making AI Less Thirsty". It calculated that GPT-3 training used 700,000 liters of fresh water at Microsoft's US data centers. The paper sparked global press coverage and led to data center water transparency demands from regulators in the EU and California.

Where data center water actually goes

Data center water consumption has three main destinations. First, on-site cooling — water that evaporates from cooling towers to reject heat from server racks. Second, off-site power generation — water consumed at the thermal power plants supplying electricity. Third, hardware manufacturing — water used to fabricate the chips, GPUs, and cooling equipment.

For a typical data center running on grid electricity, the split is roughly 20% on-site cooling and 80% off-site power generation. Renewable-energy-powered data centers shift the balance — solar and wind use almost no water (0.0 L/kWh), so the on-site cooling becomes the dominant component. Nuclear plants use 1.5–3 L/kWh, similar to coal and gas.

  • On-site cooling = 20% of AI water (evaporative cooling towers)
  • Off-site power generation = 80% of AI water (thermal power plants)
  • Hardware manufacturing = 1–3 L per CPU/GPU (low fraction of total)
  • WUE industry average = 1.9 L/kWh (Green Grid 2024)
  • Google's best data centers = 0.07 L/kWh (Finland, free-air cooling)
  • Microsoft fleet average WUE = 0.49 L/kWh (2023 report)
  • Arizona/Nevada data centers = 4.0+ L/kWh (arid summer cooling)
  • Northern Europe data centers = 0.3–0.5 L/kWh (free cooling available)

AI water footprint per ChatGPT query

A single ChatGPT query uses somewhere between 0.3 mL and 25 mL of water depending on whose number you believe. UC Riverside's 2023 paper measured GPT-3 at 519 mL per query (assuming 100 tokens and a US Iowa data center in July). OpenAI's CEO Sam Altman in 2025 cited 0.3 mL per query. Google's 2025 estimate for Gemini is 0.26 mL per query. Independent estimates from 2024 put the figure at 5 to 15 mL.

The wide range comes from different assumptions: which data center, which season, how many tokens per query, whether to include off-site power generation, which model class. Older models like GPT-3 used more energy per token. Newer models (GPT-4 Turbo, Claude 3.5, Gemini 1.5 Pro) are more efficient. Larger context windows mean more tokens per query, which scales water use up. Northern data centers in mild climates use less water than southern data centers in summer.

Training versus inference water cost

Training a large language model uses water once but at enormous scale. UC Riverside estimated 700,000 liters to train GPT-3 — roughly 28 average American households' annual water consumption. GPT-4 is estimated at 1.5 million liters, though OpenAI has not published exact numbers. LLaMA-65B (Meta, 2023) used about 80,000 liters because Meta's data centers have better cooling efficiency.

Inference is the per-query cost. It happens millions of times per day per model. At 5 mL per ChatGPT query and 200 million daily active users sending an average of 5 queries each, ChatGPT alone consumes about 5 million liters of water per day — 2 Olympic swimming pools. Multiply that by 365 days, multiply again by the number of AI services worldwide, and the global figure climbs into billions of liters per year.

Train GPT-3 once
700,000 L
≈ 0.28 Olympic pools
ChatGPT daily inference
~5,000,000 L
≈ 2 Olympic pools per day

Cooling methods and water tradeoffs

Three main cooling strategies dominate modern data centers. Water cooling (evaporative cooling towers) is the most common — moves heat efficiently but evaporates water. Air cooling (massive fans pushing outside air through server racks) uses minimal water but consumes more electricity for the fans, which indirectly uses more water at the power plant. Immersion cooling submerges chips in non-conductive fluid like 3M Novec — cuts on-site water by 95% but costs about 3x more upfront.

The choice depends on location. Hot, dry climates (Arizona, Nevada, Texas) tend toward water cooling because air cooling is too expensive when the outdoor temperature exceeds the server-rack temperature. Cool climates (Iceland, Norway, Finland) can use free-air cooling almost year-round, with WUE under 0.3 L/kWh. The push for AI compute in arid southwestern states puts pressure on water supplies that are already stressed by drought.

The siting paradox

Tech companies have built data centers in Arizona, Texas, and Saudi Arabia — regions with cheap power and weak water regulations but the hottest summers. Cooling water demands in these locations exceed those in northern locations by 3 to 10 times. The 2022 Arizona drought forced one Microsoft data center to pause expansion plans, and Google paused construction near The Dalles, Oregon over water-permit conflicts.

Global scale of AI water consumption

US data center water consumption grew from 21.2 billion liters in 2014 to about 66 billion liters in 2023, an increase of 211% over nine years. Most of that growth is attributable to AI workloads, especially after the 2023 ChatGPT boom. The trend continues — 2024 estimates put US data center water at ~80 billion liters.

Projections by industry researchers (Trends Research & Advisory, 2024) put global AI water demand at 4.2 to 6.6 billion cubic meters per year by 2027. That equals the annual water consumption of Denmark or about half the United Kingdom. Even the lower estimate puts AI water demand at 1% of global freshwater extraction, comparable to the entire concrete industry.

How to reduce your AI water footprint

Individual users have limited direct control, but a few choices matter. First, use smaller models when the task allows — Gemini Flash, Claude Haiku, and GPT-3.5-Turbo use roughly 10–20% the water of large frontier models. Second, batch your queries — one well-formed query with full context uses less water than five back-and-forth exchanges. Third, choose cloud providers that publish their WUE — Google Cloud (0.95 L/kWh) and Microsoft Azure (0.49 L/kWh) report transparently; many smaller providers do not.

Institutional pressure works better. Microsoft and Google have committed to "water positive" operations by 2030, meaning they will return more water to local watersheds than they consume. OpenAI has not made an equivalent commitment as of 2025. The EU Sustainability Reporting Directive (CSRD) may require water disclosure by 2026; the US has no equivalent federal requirement yet.

Tip

If you run AI inference at scale (B2B software, research, or large-volume content generation), pick a region with low WUE. Google's Finland (Hamina) data center has WUE near 0.07 L/kWh. Microsoft's Sweden region is similar. By contrast, US-South-Central regions run 2–4 L/kWh in summer. The carbon-aware computing literature applies equally to water-aware computing.

Controversy over the per-query water number

The "AI water footprint" became a public flashpoint after UC Riverside's 2023 paper. Critics (notably the Sean Goedecke blog post and OpenAI's own statements) argued that the 519 mL figure assumed an inefficient data center, peak summer conditions, and outdated GPT-3 architecture — not representative of modern AI. OpenAI's 2025 claim of 0.3 mL per query is 1700 times lower, which is hard to verify because OpenAI has not published its methodology.

The truth is methodological. UC Riverside used a transparent accounting that included off-site power generation water. OpenAI's lower figure likely excludes off-site water (counting only direct data-center consumption) and uses newer, more efficient architectures. Both numbers are arguably correct under their respective accounting frameworks. For consumer-facing reporting, the mid-range estimate of 1–10 mL per query is more defensible than either extreme.

AI water footprint cheat sheet
1 query (GPT-4) ≈ 5 mL water
100 queries ≈ 0.5 L = 1 bottle
10,000 queries ≈ 50 L = 0.5 shower
1,000,000 queries ≈ 5000 L (about 50 bathtubs)
Train GPT-3 = 700,000 L = 28 US households/yr

FAQ

Estimates range from 0.3 mL to 500 mL per query. UC Riverside’s 2023 paper estimated 519 mL for GPT-3 (100 tokens). OpenAI’s 2025 figure is 0.3 mL. Our calculator uses 5 mL as a mid-range estimate for GPT-4-class models, reflecting newer efficiency improvements but staying conservative.
Server racks generate heat that has to be removed. Most data centers cool the air around servers with chilled water, and that water evaporates in cooling towers. About 80% of the water consumption comes from power plants (which also need water for cooling) that generate the electricity, not from the data center directly.
About 700,000 liters of fresh water, according to UC Riverside (2023). That is equivalent to the annual freshwater use of ~28 average Americans. Training GPT-4 is estimated at around 1.5 million liters, though OpenAI has not published exact numbers.
It is growing fast. US data center water consumption tripled from 21.2 billion liters in 2014 to ~66 billion liters in 2023. Projections for global AI by 2027 reach 4.2–6.6 billion cubic meters per year — comparable to the entire water consumption of Denmark or half the UK.
Largely, yes. Immersion cooling submerges chips in a non-conductive fluid (3M Novec or similar), cutting on-site water use by about 95% (WUE drops from 2–5 L/kWh to ~0.1 L/kWh). It is still expensive (~3x cost) and not widely deployed, but Google and Microsoft have pilot installations.
A 10-minute shower uses about 100 liters. At 5 mL per query, you would need 20,000 ChatGPT queries to equal one shower. At UC Riverside’s older 500 mL estimate, you only need 200 queries — but that figure is now considered an upper bound from 2023.
In the US: data centers in Arizona, Nevada, and Texas use the most water because they are in arid regions with hot summers requiring intensive cooling. Globally, China and the southern US are the highest-consumption regions. Northern data centers in Iceland and Norway use much less.
A few options: (1) use smaller models when possible (Gemini Flash, Claude Haiku use ~10% the water of large models), (2) batch queries instead of running 10 separate ones, (3) choose cloud providers with low WUE — Google Cloud and Microsoft Azure publish theirs (0.95 and 0.49 L/kWh respectively in 2023), and (4) avoid running inference in arid regions.