Methodology & Sources

Transparency is key. Here is how we estimate the numbers.

Core Assumptions

Estimating the exact footprint of proprietary closed-source models (like GPT-4) is difficult because companies do not release exact energy data per query. However, using research papers and industry benchmarks, we can create high-confidence estimates.

⚡

Energy per Request

  • GPT-4:~0.003 - 0.01 kWh / request
  • GPT-3.5:~0.0003 kWh / request
  • Images:~0.012 kWh / image (Smartphone charge)
🌍

Carbon Intensity

We convert Energy (kWh) to Carbon (gCO2e) based on grid intensity:

  • Global Avg:~475 gCO2e/kWh
  • Green Grid:~20 gCO2e/kWh (Hydro/Wind)
💧

Water Consumption

Data centers consume water for cooling (Scope 1) and electricity generation (Scope 2). Research indicates a significant "water footprint" for AI.

â„šī¸
Key Benchmark

"Making AI Less Thirsty" (Li et al., 2023) estimates that a conversation of 20-50 questions consumes ~500ml of water.

Data Sources & references

Making AI Less Thirsty

Li et al., 2023 (University of California, Riverside)

Read Paper →
OECD AI Policy Observatory

Global policy guidance on AI computation footprint.

View Report →
Corporate Sustainability Report

Microsoft Sustainability Report (2024)

Microsoft Report →
Corporate Sustainability Report

Google Sustainability Report (2025)

Google Report →
Hugging Face & CMU

Power Hungry Processing (Luccioni et al.)

Read Paper →

Disclaimer: These figures are estimates intended for educational purposes. The AI landscape changes daily. We update our constants as new peer-reviewed data becomes available.