AI carbon footprint in EU North - Sweden: emissions per token and per query
Factor version soma-ai-ef-2026.10-dataset-v3.0.0. Last updated . Location-based grid averages, not renewable contracts. Estimates from published benchmarks.
Part of the SOMA AI Emission Factors cluster.
How much CO₂ does an AI query emit when served from EU North - Sweden?
A 1210-token query to a class B model (gpt-4o) served from EU North - Sweden emits about 0.0094 g CO₂e, and a million tokens emit 0.00714 kg CO₂e.
Rank 1 of 16 for class B, lowest carbon first (full table). This is the lowest region, and Asia Pacific - Singapore is 5.6x higher.
Which grid supplies the electricity in EU North - Sweden?
EU North - Sweden maps to Sweden, a national grid, at 38.4 g CO₂e per kWh for 2023. That is 0.2 times the EU-27 average (236 g) and 0.1 times the global average (483.2 g). The source is Ember yearly electricity data, CO2 intensity 2023, 38.42 g/kWh (lifecycle, IPCC AR5).
Cloud regions on this grid, 3 in all (1 AWS, 1 Azure, 1 Google Cloud).
| Provider | Region id | Region name | Location | Mapping confidence | Mapping note |
|---|---|---|---|---|---|
| AWS | eu-north-1 | Europe (Stockholm) | Stockholm | high | |
| Azure | swedencentral | Sweden Central | Gävle | high | |
| Google Cloud | europe-north2 | europe-north2 | Stockholm, Sweden | high |
Which grids have an intensity close to Sweden?
| Difference g | Grid | Code | g CO₂e per kWh | Ratio |
|---|---|---|---|---|
| -3.6 | Switzerland | CHE | 34.8 | 0.906 |
| -7.9 | Norway | NOR | 30.5 | 0.794 |
| 9.2 | Namibia | NAM | 47.6 | 1.240 |
| 9.2 | Sierra Leone | SLE | 47.6 | 1.240 |
| -10.6 | Iceland | ISL | 27.8 | 0.724 |
| -11.5 | Congo (the Democratic Republic of the) | COD | 26.9 | 0.701 |
| -11.9 | Quebec (Canada) | CAN-QC | 26.5 | 0.690 |
| -13.7 | Costa Rica | CRI | 24.7 | 0.643 |
| -13.9 | Paraguay | PRY | 24.5 | 0.638 |
| -14.2 | Bhutan | BTN | 24.2 | 0.630 |
Emission factors for EU North - Sweden by model class
| Model class | Electricity kg | Hardware kg | Construction kg | Training kg | Total kg CO₂e | g per query | Water L | mL per query |
|---|---|---|---|---|---|---|---|---|
| Class A (small) | 0.00084 | 0.00027 | 0.00014 | 0.00060 | 0.002 | 0.0025 | 0.131 | 0.18 |
| Class B (mid-size) | 0.00274 | 0.00101 | 0.00046 | 0.00293 | 0.007 | 0.0094 | 0.430 | 0.59 |
| Class C (frontier) | 0.01020 | 0.00388 | 0.00169 | 0.01284 | 0.029 | 0.037 | 1.601 | 2.2 |
| Model class | Central | Class range low | Class range high | High case | B200 | Development |
|---|---|---|---|---|---|---|
| Class A | 0.001848 | 0.001331 | 0.002455 | 0.002511 | 0.001714 | 0.001141 |
| Class B | 0.00714 | 0.005472 | 0.008408 | 0.009346 | 0.006065 | 0.003735 |
| Class C | 0.02862 | 0.01834 | 0.03672 | 0.03687 | 0.02899 | 0.01391 |
How much does siting matter for a model served from EU North - Sweden?
Electricity is 38.4% of the class B total here (class A 45.3%, class C 35.6%), with hardware at 14%, construction 6.4% and training 41%. A cleaner grid shrinks the electricity term, while the other 61.6% stays put.
No other region in the dataset has a lower class B total than the 0.00714 kg per million tokens printed here, so there is nothing to gain by moving.
How much water does AI usage consume in EU North - Sweden?
A class B query of 1210 tokens from EU North - Sweden consumes about 0.59 mL of water and a million tokens 0.43 litres, or 0.0996 litres without the water attributed to hydropower.
| Input | Value |
|---|---|
| On-site WUE, L per kWh | 0.02 |
| WUE denominator | IT load |
| WUE basis as published | withdrawal |
| Withdrawal to consumption | 0.6667 |
| EWIF, L per kWh | 6.019 |
| EWIF without hydro, L per kWh | 1.386 |
| AWARE factor | 2.17 |
| Confidence | published-site |
| Source | AWS; AWS Region Europe (Stockholm) eu-north-1; 2025 |
| Model class | On site | Grid electricity | Total | Total without hydro | AWARE-weighted on site |
|---|---|---|---|---|---|
| Class A | 0.000242 | 0.131 | 0.131 | 0.0304 | 0.000525 |
| Class B | 0.000792 | 0.429 | 0.43 | 0.0996 | 0.00172 |
| Class C | 0.00295 | 1.6 | 1.6 | 0.371 | 0.0064 |
Where do these numbers come from and how do I cite them?
Figures per query are one plain exchange, before the Tier 2b reasoning uplift. Electricity term. GPU energy per token from the ML.ENERGY Leaderboard v3, a facility overhead of 1.20, and grid carbon intensity from EPA eGRID 2023 and Ember 2023. Embodied hardware, data-centre construction and amortised training terms are documented on the methodology page. Water combines on-site cooling water (Li et al. 2025) with the water embedded in grid electricity (Reig et al. 2020). The modelled range of the models inside a class, around the class value, is −41% to +48% for class A, −39% to +24% for class B and −67% to +55% for class C.
Dataset: AI Inference Emission and Resource Factors for Corporate GHG Inventories (SOMA AI Emission Factors (SAEF)), Llopis-Garcia, G., SOMA AI, 19 October 2026, CC BY 4.0, concept DOI 10.5281/zenodo.20443585, which resolves to the latest version (this version 10.5281/zenodo.23371447, earlier version 2 values at 10.5281/zenodo.22767475). Method: Accounting for Purchased AI Inference in Corporate GHG Inventories: A Token-Based Method, arXiv:2606.10660, under peer review.
Partners can fetch any row from the AI factors API, and the whole table is a CSV download.
Cite as "0.00714 kg CO₂e per million tokens, class B, EU North - Sweden, SAEF, 19 October 2026" and link somaai.earth/ai-footprint/factors/eu-sweden.