Image Generation

GPT-4o Image Generation Environmental Impact

HeavyEstimated

Native image generation inside ChatGPT

Architecture
Autoregressive image generation (native to GPT-4o)
Provider
OpenAI
4.5 Wh
Energy per query
2.0 g
CO₂ per query
16 mL
Water per query
15x more than
vs Google search

Energy per query

4.5 Wh

15x more than a Google search (0.3 Wh)

CO2 per query

2.0 g

US East (Virginia) grid (450 gCO₂/kWh)

Water per query

16 mL

~63 queries to fill 1 litre

Processing location

Azure US East / Sweden

Provider

OpenAI

Category

Image Generation

Grid carbon intensity

450 g CO2/kWh (25% renewable)

How does GPT-4o Image Generation compare?

Ranked #109 of 152 models by energy per query

0 Wh2 Wh4 Wh6 Wh8 WhStable DiffusionXLDALL-E 3GPT-4o ImageGenerationMidjourneyGoogle search (0.3 Wh)

Detailed Breakdown

Energy Consumption

GPT-4o's native image generation is distinct from DALL-E 3 — it uses autoregressive token generation rather than a separate diffusion model. Estimated at ~4.5 Wh per image, reflecting the full GPT-4o model generating high-resolution image tokens alongside text reasoning. This is more expensive than DALL-E 3 alone because the entire GPT-4o model is engaged.

Power Source & Carbon

Runs on Microsoft Azure, sharing infrastructure with GPT-4o text inference.

Water Usage

At approximately 16 mL per image, reflecting the extended GPU time required for autoregressive image token generation.

About GPT-4o Image Generation

GPT-4o Image Generation is a image generation model from OpenAI, released in March 25, 2025. Native image generation inside ChatGPT. Each query uses 4.5 Wh of energy and produces 2.0 g of CO₂. That's 15x the energy of a Google search — reflecting the computational demands of image generation.

These figures are estimates derived from hardware specifications and API benchmarks — OpenAI has not published official energy data for GPT-4o Image Generation. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.

GPT-4o Image Generation in Context

450 MWh
estimated daily

At global scale

With an estimated 10M+ daily users averaging 10 queries each, GPT-4o Image Generation consumes roughly 450 MWh of electricity per day — enough to power 15000 homes.

41.1 kWh
per year

Your yearly GPT-4o Image Generation footprint

At 25 queries per day, your annual GPT-4o Image Generation usage consumes 41.1 kWh — roughly what a fridge uses in a month. That produces 18.3 kg of CO₂.

15x
vs a text query

Image vs text: the energy gap

Generating one image with GPT-4o Image Generation uses 15x more energy than a text conversation with ChatGPT. The diffusion process that creates images requires many iterative passes through the model, each refining the image from noise — far more compute per output than generating text one token at a time.

What does your GPT-4o Image Generation usage cost the planet?

Use our calculator to estimate your personal environmental footprint based on how often you use GPT-4o Image Generation.

Calculate My Compute

Frequently Asked Questions

How much energy does GPT-4o Image Generation use per query?

Each GPT-4o Image Generation query consumes approximately 4.5 Wh of energy. This is 15x more than a traditional Google search (~0.3 Wh).

What is GPT-4o Image Generation's carbon footprint?

Based on the carbon intensity of Azure US East / Sweden, each query produces approximately 2.0 g of CO2. The grid in this region has a carbon intensity of 450 g CO2/kWh with 25% renewable energy.

How much water does GPT-4o Image Generation use?

Each query consumes approximately 16 mL of water, primarily used for cooling the data centers that process the request.

How does GPT-4o Image Generation compare to a Google search?

A GPT-4o Image Generation query uses 15x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while GPT-4o Image Generation uses 4.5 Wh.

Technical Details

Architecture

Autoregressive image generation (native to GPT-4o)

Release date

2025-03-25

Open source

No