SOMA AI Emission Factors changelog

Every published version of the dataset, newest first, so an inventory can name the version it used. Current: Factor version soma-ai-ef-2026.11-dataset-v3.1.0. Factors last updated .

Page updated

Part of the SOMA AI Emission Factors cluster.

How are the factors versioned?

Each release of the SOMA AI Emission Factors (SAEF) gets a Zenodo record and a new version. The concept DOI 10.5281/zenodo.20443585 always resolves to the latest one, and each record DOI below resolves to exactly that version. SOMA's own factor version string, currently soma-ai-ef-2026.11-dataset-v3.1.0, names the copy of the dataset behind every figure on this site and in the API, so quote it beside the DOI when a reader has to reproduce a number. A published version never changes: a new release gets a new version.

Version history

  1. Version 3.1.0, published

    Record 10.5281/zenodo.24160409. SOMA factor version soma-ai-ef-2026.11-dataset-v3.1.0.

    • Azure canadaeast and Google Cloud northamerica-northeast1, which the providers place in Quebec, now take the Quebec provincial grid (0.0265 kg CO2e/kWh) instead of the Canadian national average (0.1744), as AWS ca-central-1 already did; their class B factor per million billed tokens falls from 0.0168 to 0.0063 kg.
    • The Central African Republic grid row, published by Ember as 0 g/kWh because its emissions round to 0.00 MtCO2, now takes a documented zero-intensity fallback: 0.0238 kg/kWh (100% hydro x the World hydro life-cycle factor).
    • No other row changes; every number printed in the method paper is unchanged.
  2. Version 3.0.0, published

    Record 10.5281/zenodo.23371447. SOMA factor version soma-ai-ef-2026.10-dataset-v3.0.0.

    • Four-term lifecycle boundary (electricity, embodied hardware, data-centre construction and training) with a separate development line.
    • Effective-token basis with input weight W_in 0.30 and a per-run prefix-cache share; 0.3025 effective tokens per billed token at the default mix.
    • Three model classes (A small, B standard, C frontier), each the mean of its benchmarked members: 18 open-weight models on H100 and B200 from the ML.ENERGY Leaderboard v3. Versions 1 and 2 used a named proxy model per class.
    • Coverage of 210 countries, 27 eGRID subregions and 143 cloud regions, with water for 17 regions using AWARE 2.0.
    • Monte Carlo uncertainty ranges.
    • The reasoning uplift for Tier 2b applies to output tokens only.
    • Headline spreads on the inference footprint: class A to C about 12x, Singapore to Sweden about 9x (electricity only about 13x).
  3. Version 2.0.0, published

    Record 10.5281/zenodo.22767475. SOMA factor version soma-ai-ef-2026.09-lifecycle-v2.

    • Added regions France (eu-west-3), Finland (europe-north1), UK (europe-west2), Belgium (europe-west1) and Iowa (us-central1).
    • Completed the carbon rows for us-west-2, us-south-central and us-south.
    • Corrected two v1 rows: us-east-1 class B carbon from 0.04380 to 0.04378 kg CO2e per M tokens, and Sweden class C water from 1.254 to 1.253 L per M tokens.
    • Introduced lifecycle add-ons (embodied hardware and training amortisation).
    • Restated the headline regional spread from 13x to 12x electricity only and about 6x with add-ons, the earlier figure being a rounding artefact.
  4. Version 1.0.0, published

    Record 10.5281/zenodo.20443586. SOMA factor version soma-ai-ef-2026.09-lifecycle-v1.

    • First release, accompanying arXiv 2606.10660 v1.
    • Four classes (small, mid, large MoE, large dense).
    • Electricity-only factors per region and class.
    • Regional spread about 13x, Singapore to Sweden.

The browsable tables for the current version are on the SOMA AI Emission Factors (SAEF) page.

Keep reading