Cheese (30 g) = 3,600 AI prompts
Cheese (30 g) equals roughly 3,600 AI prompts on the CO₂ side: 720 g, while an average text prompt weighs just 0.2 g. The math, the source and detailed equivalents are right below, with public, verifiable numbers.
Figures updated on June 30, 2026 · Source : Our World in Data : GES par aliment (Poore & Nemecek 2018)
The CO₂ of a prompt reflects the emissions of the electricity used while answering, on an average power grid. On the action side, figures include production (food, goods) or combustion (transport). Here again, we compare orders of magnitude, not gram-perfect measurements.
Why 3,600 prompts?
The math fits in three lines, using public, sourced values:
- 1.Cheese (30 g): 720 g (Our World in Data : GES par aliment (Poore & Nemecek 2018))
- 2.1 text prompt: 0.2 g (published range: 0.15 to 1 g)
- 3.720 g ÷ 0.2 g ≈ 3,600 prompts
In other words, “Cheese (30 g)” weighs as much as 3,600 questions asked to a generative AI. It is an order of magnitude, not a to-the-decimal measurement: sources publish ranges, and we use a median value, explained in our methodology.
What about a generated image or a reasoning prompt?
Not every AI use weighs the same. A generated image costs about 6 g and a “reasoning” prompt about 13.3 g, versus 0.2 g for a simple text prompt. So “Cheese (30 g)” is also roughly 120 times one generated image, or 54 times one reasoning prompt.
Convert your own usageWhere do these numbers come from?
The value we use for “Cheese (30 g)” comes from “Our World in Data : GES par aliment (Poore & Nemecek 2018)”, referenced like all our data on the Sources page. No number on this page is made up: every value can be traced back to a public document.
On the prompt side, there is no official meter: we convert the electricity used with the average grid intensity (about 480 g of CO₂ per kWh), which lands around 0.15 g in operation. We use 0.2 g per prompt, rising towards 1 g over the full life cycle.