Standard text queries
Quick questions, rewriting, brainstorming, and everyday chat.
Estimate, compare, decide
Worth the Compute?
See the real-world footprint of how you use AI.
AI has an environmental footprint, but the impact of one quick question is very different from generating hundreds of images or videos. Build a scenario to put your usage into perspective.
Directional estimates based on public research, not exact meter readings.
Your AI usage
Monthly usage
Your AI usage
Estimate your monthly AI usage
Ask your AI to help build an estimate, then paste the result here.
- 1. Copy the prompt
- 2. Answer a few questions in your AI assistant
- 3. Paste the estimate back here
Works with ChatGPT, Claude, Gemini, and other AI assistants.
Prefer to enter your usage manually?
Add or adjust text queries, coding sessions, images, automations, and other AI activity yourself.
Estimated impact
Your approximate monthly footprint
Based on the monthly usage scenario shown.
Water
3.8 L
Operational water estimate
Carbon
0.8 kg CO2e
Grid-average estimate
Energy
2.1 kWh
0.4-3.8 kWh range
Actual impact varies by model, hardware, data center, region, prompt length, response length, and whether renewable energy or water-efficient cooling is used.
Activity breakdown
Standard text queries0.5 kWh
Complex text queries1 kWh
Image generation0.6 kWh
Methodology
How this MVP estimates impact
Each activity is normalized into one typical month, then uses a low-to-high energy range per request or generation.
Carbon uses a broad grid-average factor of 0.39 kg CO2e per kWh. Water uses 1.8 liters per kWh for operational consumption.
Equivalencies use centralized modeled assumptions for familiar activities such as charging devices, driving, showers, laundry, and tree carbon absorption.
Agent and automation runs are currently modeled as deep reasoning tasks because this MVP does not yet have a distinct automation energy factor.