Skip to content
by Projet Celsius
Projet Celsius
Key figureUsesUpdated on 11 November 2026

2.9Wh

Electricity use of an AI-generated image

Average of the open models measured · one image · study published in 2024

Generating an image with an open model consumes 2.9 Wh of electricity on average, according to the measurements by Luccioni, Jernite and Strubell published in 2024; the most energy-hungry model in the study reaches 11.49 Wh per image.

An average: the median generator consumes 1.35 Wh, the most energy-hungry 11.49 Wh.

See the source
A dismantled smartphone on a repair mat, next to its battery
A dismantled smartphone, next to its battery. Photo Fotografia Lui Vlad, Pexels

Definition and scope

Average electricity consumed to generate an image, measured on the open image generation models tested by the study "Power Hungry Processing".

What is included

  • The GPU during generation

What is excluded

  • The rest of the server and the cooling
  • Training and manufacturing
  • The user's device

How the value is obtained

Measurement of the GPU's electricity over 1,000 generations, for each open model tested, then average of the image generation models.

Limits and uncertainties

  • Open models from 2023; no provider of a commercial generator publishes a measurement.
  • The image resolution and the number of computation steps change the result a great deal.

The order of magnitude

All figures

2.9 Wh on an electricity scale, with 6 other reference figures, brought to the same unit. Each tick multiplies by 10.

1 Wh1 kWh1 MWh1 GWh1 TWh1,000 TWh2.9 WhAI-generated image0.24 WhText query to Gemini0.3 WhText question to an AI90 WhGenerated 5-second video100 to 174 TWhAI worldwide in 2025415 TWhData centres in 2024945 TWhData centres in 2030
1 Wh1 kWh1 MWh1 GWh1 TWh1,000 TWh945 TWhData centres in 2030415 TWhData centres in 2024100 to 174 TWhAI worldwide in 202590 WhGenerated 5-second video2.9 WhAI-generated image0.3 WhText question to an AI0.24 WhText query to Gemini
This figureAnother reference figure (link to its entry)Logarithmic scale

In context

See the chart and its data

Electricity use by type of AI task

For 1,000 inferences, in kilowatt-hours (logarithmic scale)

Measurements published in 2024

Generating costs far more than classifying: image generation consumes more than 1,000 times as much as text classification.

0.0010.01× 100.1× 101× 1010× 10ScalelogarithmickWh (logarithmic scale)Image generation2.9Text summarisation0.049Text generation0.047Object detection0.038Image classification0.007Text classification0.002
Source: Luccioni et al., Power Hungry Processing, 2024 · projetcelsius.com/en/ai · CC BY 4.0AI Impactby Projet Celsius

Update history

The hub changelog
Answer sentence added at the top of the entry
First version of the entry
Value set for the series of articles on AI

More on uses

The whole Uses topic →

Charts and maps

ChartUsesElectricity by type of AI taskover1,000timesimage generation versus text classificationChartUsesReasoning mode multiplies energy useChartUsesEnergy savings from 4 settings10times as muchsaved by shortening the response rather than the instructionChartUsesElectricity used by a question, an image, a videoabout300timesa 5-second video versus one question

Analyses

Where does this figure come from?Uses20 October 2026"A generated image is a phone charge"The comparison did the rounds of the media in late 2023. It came from a serious study, but from its preliminary version, from the most power-hungry generator, and from a phone charge counted as 12 Wh.Where does this figure come from?Uses20 October 2026"A generated video uses 1 kWh"The figure has been circulating since 2025, often attached to Sora. It comes from a real measurement, but one made on a different, open model, in a particularly heavy version. Measurements of 7 generators range from 0.14 to 415 Wh.

Tools

ToolUsesEstimate the footprint of your AI use3kgCO2eequivalent to 21 km by combustion-engine car

On the Projet Celsius blog ↗

Blog article ↗UsesWhat is the carbon footprint of an AI-generated image?

Across the hub

AI Impact DatabaseAll the figures, their methodology and their sources.AI Ranking10 assistants rated out of 100: performance, ethics and environment.AnalysesThe signed pieces: analyses, viewpoints, "Where does this figure come from?".Charts and mapsYou are hereThe reference visuals, with their data and sources.ToolsOpen calculations, your assumptions on the left, the result on the right.

A question about a figure?

An error, a source to add, a methodology question, a press request: write to us. Guillaume Pakula, Clément Reynaud, Sébastien Pierfederici and Alexandre Veilly will reply.

Write to us →

Projet Celsius, which runs this hub, also helps companies measure and reduce the footprint of their AI use.

Accounting for AI in your Bilan Carbone® ↗
Projet Celsius, environmental strategy consultancy

We support companies that account for AI in their Bilan Carbone®

Visiter Projet Celsius ↗
  • Institut Pasteur
  • SNCF Gares & Connexions
  • Radio France
  • Institut Curie
  • Institut Polytechnique de Paris
  • ESSEC Business School
  • Qualiopi, certified process