AI Footprint Calculator

Impact

One typical AI query + response for the Use Case

0.648 Wh
Energy
0.173 g CO₂e
Carbon
3.25 mL
Water

Impact Over Time

⚡ Energy141 Wh
🫖 kettle boilings0.7 boilings
☁️ Carbon37.7 g CO₂e
🍌🚶 banana-powered walk0.9 km
💧 Water708 mL
🍼 water bottles1 bottles

% of Londoner Monthly Usage

⚡ Electricity consumption0.0401%
🌍 Carbon footprint0.0035%
💧 Water consumption0.0162%
Per person UK average: 351 kWh electricity, 1,083 kg CO₂e, 4,380 L water

ChatGPT Monthly Carbon Emissions by Data Center

Configuration:
Location-Based (actual grid mix)
Market-Based (with renewable procurement)
A
B

Configuration Comparison

🔄 Key Differences:
Region: 🇺🇸 eastus2 vs 🌍 swedencentral
Energy
-0.0%
A: 141 Wh
B: 141 Wh
Carbon
-80.8%
A: 37.7 g CO₂e
B: 7.22 g CO₂e
Water
-0.0%
A: 708 mL
B: 708 mL

Estimate the environmental impact of your AI usage

Based on Google's comprehensive production measurements (2024) and EPA eGRID 2022/Green-e 2024 residual data. Energy model calibrated for GPT-4o baseline, with median use case targeting ~0.03g CO2e (market-based). Includes full system overhead: AI accelerators, CPU/DRAM, idle capacity, and data center PUE.

Location-based carbon uses regional grid intensity. Market-based carbon reflects renewable energy procurement: GCP uses actual CFE% per region, Azure/AWS use residual-anchored heuristic.

Methodology references: Google AI Scale Study | Epoch AI Analysis | ClimateAligned Research

Model availability: GCP | Azure | AWS