Version 2026-07.1

Methodology

How we estimate the environmental cost of AI inference, and how we translate it into a donation amount. Every constant is cited; every formula is auditable.

The formula

kWh = (output_tokens + input_tokens × 0.1) × Wh/token × PUE / 1000

kg CO₂e = kWh × grid_intensity

liters H₂O = kWh × water_intensity

damage_$ = (kg CO₂e / 1000) × social_cost_of_carbon

donation = damage_$ × multiplier

Constants

ConstantValueSource
PUE (data center overhead)1.2Uptime Institute Global Survey 2023
Grid intensity0.38 kg CO₂e/kWhUS EPA eGRID national average 2022
Water intensity0.55 L/kWhLi et al. 2023, "Making AI Less Thirsty"
Social cost of carbon$190/t CO₂eUS EPA 2023 SC-GHG report, central estimate
Input token energy fraction0.1×Prefill ≪ decode; Luccioni et al. 2024

Energy per token by model class

ClassWh / output tokenBlended $/MTok (for spend fallback)
small0.0002$0.6
medium0.0006$5
large0.0025$15
frontier0.006$40

Limitations

  • These are estimates, not measurements. Actual energy use varies by data center location, cooling method, utilization, and hardware generation.
  • We use a US-average grid intensity. Many large providers use significant renewable energy, which would reduce real-world emissions.
  • Subscription tier estimates are derived from published average-usage studies and are inherently approximate.
  • The social cost of carbon is contested. $190/t is the EPA 2023 central estimate; ranges from ~$50 to $400+ depending on discount rate and model.
  • The default 2× multiplier is designed to overshoot under any reasonable uncertainty range in the underlying estimates. We suggest the donation amount; you choose whether to pay it.

Citations

  1. Luccioni, A., Jernite, Y., & Strubell, E. (2024). Power Hungry Processing: Watts Driving the Cost of AI Deployment? ACM FAccT 2024.
  2. Epoch AI (2025). How much energy does ChatGPT use?
  3. US EPA (2023). Report on the Social Cost of Greenhouse Gases.
  4. US EPA eGRID (2022). Emissions & Generation Resource Integrated Database.
  5. Uptime Institute (2023). Global Data Center Survey.
  6. Li, P. et al. (2023). Making AI Less "Thirsty": Uncovering and Addressing the Secret Water Footprint of AI Models.