Simon Willison’s Weblog

Subscribe

Our contribution to a global environmental standard for AI (via) Mistral have released environmental impact numbers for their largest model, Mistral Large 2, in more detail than I have seen from any of the other large AI labs.

The methodology sounds robust:

[...] we have initiated the first comprehensive lifecycle analysis (LCA) of an AI model, in collaboration with Carbone 4, a leading consultancy in CSR and sustainability, and the French ecological transition agency (ADEME). To ensure robustness, this study was also peer-reviewed by Resilio and Hubblo, two consultancies specializing in environmental audits in the digital industry.

Their headline numbers:

  • the environmental footprint of training Mistral Large 2: as of January 2025, and after 18 months of usage, Large 2 generated the following impacts: 
    • 20,4 ktCO₂e, 
    • 281 000 m3 of water consumed, 
    • and 660 kg Sb eq (standard unit for resource depletion). 
  • the marginal impacts of inference, more precisely the use of our AI assistant Le Chat for a 400-token response - excluding users' terminals:
    • 1.14 gCO₂e, 
    • 45 mL of water, 
    • and 0.16 mg of Sb eq.

They also published this breakdown of how the energy, water and resources were shared between different parts of the process:

Infographic showing AI system lifecycle environmental impacts across 7 stages: 1. Model conception (Download and storage of training data, developers' laptops embodied impacts and power consumption) - GHG Emissions <1%, Water Consumption <1%, Materials Consumption <1%; 2. Datacenter construction (Building and support equipment manufacturing) - <1%, <1%, 1.5%; 3. Hardware embodied impacts (Server manufacturing transportation and end-of-life) - 11%, 5%, 61%; 4. Model training & inference (Power and water use of servers and support equipment) - 85.5%, 91%, 29%; 5. Network traffic of tokens (Transfer of requests to inference clusters and responses back to users) - <1%, <1%, <1%; 6. End-user equipment (Embodied impacts and power consumption) - 3%, 2%, 7%; 7. Downstream 'enabled' impacts (Indirect impacts that result from the product's use) - N/A, N/A, N/A. Stages are grouped into Infrastructure, Computing, and Usage phases.

It's a little frustrating that "Model training & inference" are bundled in the same number (85.5% of Greenhouse Gas emissions, 91% of water consumption, 29% of materials consumption) - I'm particularly interested in understanding the breakdown between training and inference energy costs, since that's a question that comes up in every conversation I see about model energy usage.

I'd really like to see these numbers presented in context - what does 20,4 ktCO₂e actually mean? I'm not environmentally sophisticated enough to attempt an estimate myself - I tried running it through o3 (at an unknown cost in terms of CO₂ for that query) which estimated ~100 London to New York flights with 350 passengers or around 5,100 US households for a year but I have little confidence in the credibility of those numbers.

Monthly briefing

Sponsor me for $10/month and get a curated email digest of the month's most important LLM developments.

Pay me to send you less!

Sponsor & subscribe