AI SustainabilityIndex
Get certified
Forest canopy fading into a data grid
AI sustainability, made measurable

The carbon cost of your AI, in one honest number.

We help UK businesses understand the real footprint of the models, servers and people behind their AI — and shows where to cut emissions without losing pace.

Score my business See the LLM index
0.24–8 Wh
Per query, depending on model
10–30×
Reasoning runs vs. a standard query
0.207
kg CO₂e per kWh — UK grid avg
Calculator

Score your AI footprint.

Pick the LLMs your team relies on, tell us about your infrastructure, and we'll model the annual energy and emissions — and what to do about them.

Inputs
Which LLMs does your team use?
Where does your workload run?

UK national grid (DESNZ 2024).

Usage mix — what % of prompts goes to each?
GPT-4o50%
Gemini Flash50%

Mix totals 100%. We normalise automatically — round numbers are fine.

Per-query energy figures are modelled, not measured — frontier vendors don't publish verified Wh/query. Results are advisory and rounded. See methodology, sources & ranges → v1.2 · reviewed 2026-06-06

Your AI Sustainability score
70
B
/ 100
Scores are rounded to the nearest 5.
AI emissions / yr
0.07 t CO₂e
Range 0.04–0.15 t
Total digital / yr
7.7 t CO₂e
AI + servers + devices · 0.207 kg/kWh
Total energy / yr
37,111 kWh
Trees to offset
~366
Indicative only
Where the kWh comes from
  • AI usage361 kWh1% of total
  • Servers18,000 kWh49% of total
  • Devices18,750 kWh51% of total
Recommendations
  • Your on-prem server ratio is high. Consolidate workloads onto renewable-powered cloud regions (e.g. AWS eu-west-1, GCP europe-west1) to lower scope-2 emissions.
  • You're in good shape. Offset residual emissions via a verified UK woodland scheme and publish your AI footprint in your sustainability report.
LLM Index

Not all models are equal.

A curated ranking of major LLMs by reported watt-hours per query. Choose smaller models for everyday tasks; reserve frontier reasoning for problems that deserve it.

Model
Wh / query
Tier
  • Gemini FlashGoogle

    Small, efficient. Runs on TPU v5e — among the lowest reported energy per query.

    0.24
    Tier A
  • Claude HaikuAnthropic

    Lightweight Claude tier, good for high-volume tasks.

    0.40
    Tier A
  • GPT-4o miniOpenAI

    Distilled model, suitable for most everyday tasks.

    0.50
    Tier A
  • Llama 3 8B (self-hosted)Meta

    Energy depends on your grid; great on renewable-powered infra.

    0.60
    Tier B
  • Mistral SmallMistral

    Efficient European-hosted option.

    0.70
    Tier B
  • Gemini 2.5 ProGoogle

    Larger reasoning model — use selectively.

    1.80
    Tier B
  • Claude SonnetAnthropic

    Mid-tier Claude, balanced quality/energy.

    2.20
    Tier C
  • GPT-4oOpenAI

    Multimodal flagship; meaningfully more energy than 4o-mini.

    2.90
    Tier C
  • Llama 3 70B (self-hosted)Meta

    Heavy self-hosted footprint unless run on green energy.

    5.50
    Tier D
  • Claude OpusAnthropic

    Top-tier reasoning. Reserve for high-value tasks only.

    6.50
    Tier D
  • GPT-5 / o-series reasoningOpenAI

    Deep-reasoning runs can use 10–30× a standard query.

    8.00
    Tier D
Use AI Wisely

When AI earns its energy — and when it doesn't.

A working guide to which tasks justify the carbon of a model call, which should stay manual, and which need a careful hand.

Use AI

Drafting emails & meeting notes

High productivity gain, tiny model is sufficient. Tier-A models only.

Use AI

Summarising documents

Saves hours of human attention; energy per summary is negligible vs. the time saved.

Use AI

Code autocomplete & refactors

Use a small model for completion; reserve large models for architecture questions.

Use AI

Customer support triage

Route first-line responses through a Tier-A model with human escalation.

Skip

Generating images for every blog post

Image generation is ~100× more energy than text. Use only when the image earns its place.

Skip

Real-time AI video for internal demos

Highest energy class. Record once or use static assets instead.

Skip

Running 'deep research' on routine questions

Reasoning loops can use 10–30× a standard query. Save for genuinely complex problems.

Skip

AI search instead of a known URL

If you know the source, go direct. AI search is convenient but wasteful as a habit.

Careful

Multi-agent workflows

Powerful, but each step compounds energy. Cap step count and prefer smaller agents.

Careful

Fine-tuning on your data

Training is energy-intensive up-front but pays back if it lets you use a smaller model.

Use AI

Voice transcription & translation

Mature, efficient, high human value.

Careful

AI-generated slide decks end-to-end

Worth it for first drafts; don't regenerate 20 versions.

Carbon-neutral AI isn't a slogan. It's a series of small, sharp choices.

We give your team the data and defaults to make those choices — every day, in every workflow.

Run your score