ChatGPT vs Gemini

This comparison uses the current IdleForest model for ChatGPT and Gemini: their category, modeled CO2 per use unit, methodology notes, key drivers, and assumptions.

Supporting comparison page

ChatGPT

150g

CO2 / HOUR

Higher emissions

Gemini

145g

CO2 / HOUR

Data-backed comparison

Summary

When comparing ChatGPT and Gemini, ChatGPT generates significantly more CO2 emissions per hour (150g) than Gemini (145g). Both applications rely on devices, networks, and server infrastructure, which all contribute to their environmental impact.

Why the gap happens

ChatGPT is modeled at 150g CO2 per unit, while Gemini is modeled at 145g, so the visible gap is 5g in the current dataset.
Both products sit in the AI category, so the difference comes from the per-product estimate and page-level methodology fields rather than a category change.
This estimate converts repeated AI interactions into an hourly footprint by combining model inference demand with the surrounding data-center overhead.
Prompt volume and response length both change inference demand.

What to act on first

Because ChatGPT is higher in the current model, start there: Consolidate related prompts into one well-scoped request.

ChatGPT is currently modeled at 5g CO2 more per unit of use than Gemini.

Deep dive into their footprints

Read the full ChatGPT report Read the full Gemini report

Comparison takeaways

ChatGPT is modeled at 150g CO2 per unit, while Gemini is modeled at 145g, so the visible gap is 5g in the current dataset.

Both products sit in the AI category, so the difference comes from the per-product estimate and page-level methodology fields rather than a category change.

About IdleForest

IdleForest is a desktop app that plants trees in the background while your computer is idle. It uses your unused internet bandwidth to fund reforestation projects.

Connect Desktop App