How much CO₂ does one ChatGPT query produce?

Estimates from public research range from 0.3 to 5 grams of CO₂e per query, depending on model size, output length and the carbon intensity of the data centre's electricity supply. The calculator returns a model-specific figure.

Why does region matter?

The same compute uses very different amounts of fossil-based electricity depending on where the data centre is. Iceland (~10 g/kWh) runs on hydro and geothermal; coal-heavy grids approach 800 g/kWh.

How accurate is this?

It's a rough order-of-magnitude estimate. Public energy-per-token figures vary by 5× depending on hardware, batching efficiency and routing. Use the result for awareness, not for compliance reporting.

🌱 AI Kulstofsaftryk Lommeregner

Estimér kWh og gram CO₂ produceret af en AI-forespørgsel — efter model og strømregion.

Model

Input-tokens

Output-tokens

Netregion (kulstofintensitet)

Prompts per day (optional)

ℹ️ Order-of-magnitude estimate. Energi-per-token figures vary by hardware generation and batching. Kulstof intensity uses 2024 averages from Ember and Climatiq.

About the AI Kulstofsaftryk Lommeregner

Every prompt sent to a hosted LLM consumes a small amount of electricity to power inference servers. Multiplied across millions of users and billions of queries, it adds up. The AI Carbon Footprint Calculator turns model + token-count + grid-region into an estimated kWh and grams of CO₂ equivalent (gCO₂e).

The formula

CO₂ = tokens × Wh-per-token × netintensitet

Energi per token by model class

Small (≈ 7B params): ~0,2 mWh / token (Llama 3 8B, Mistral 7B)
Medium (≈ 70B): ~1 mWh / token (Llama 3 70B, Claude Haiku)
Large (≈ 400B+): ~3 mWh / token (GPT-4o, Claude Sonnet)
Frontier (≈ 1T+): ~6 mWh / token (GPT-4 / Claude Opus klasse)

Output-tokens koster mere end input-tokens (typisk 2–3× fordi hvert output kræver et fremadgående gennemløb). Tallene er omtrentlige — estimater fra 2024–2025 varierer med 3–5×.

Kulstof intensity by region (gCO₂e / kWh)

Island: 11 (geotermisk + vandkraft)
France: 56 (nuclear)
UK: 230
EU gennemsnit: 250
Tyskland: 380
US gennemsnit: 387
Texas (ERCOT): 410
India: 720
Coal-heavy grid: ~820

Ofte stillede spørgsmål

Is this an exact number? No — treat it as an order-of-magnitude estimate. Real-world variation depends on hardware, batching efficiency, model routing and exact regional carbon intensity at time of inference.

What about model training? Træningsemissioner amortiseres over en models levetid. For bredt anvendte modeller bidrager træning med ca. 5–20% af de samlede livscyklusemissioner.

How does this compare to a Google search? A standard Google search produces about 0.2 g CO₂e. A short LLM prompt to a frontier model is roughly 5–25× higher.

Par dette med vores AI Model Cost Calculator og Prompt Cost Calculator for et fuldt bæredygtigheds- og omkostningsbillede.

⚠️

Vigtig bemærkning: Energi-per-token og kulstofintensitetstal ændrer sig hurtigt. Brug dette værktøj til bevidsthed; konsulter udgiverens bæredygtighedsrapporter for overholdelsesrapportering.

Relaterede lommeregnere

AI Model Cost Calculator — Sammenlign API-priser for 100+ LLM'er inkl. GPT-4o, Claude, Gemini og Llama. Beregn præcise tokenomkostninger.
AI vs Human ROI Calculator — Sammenlign AI-værktøjsomkostninger med menneskelig arbejdskraft. Månedlige besparelser, årligt ROI og break-even med live modelpriser.
Carbon Footprint Calculator — Estimat your annual CO₂ emissions from transport, home energy, diet and lifestyle. Sammenlign to global averages.