9.0× less energy than Mistral Large 2.
Mistral Large 2 Environmental Impact
Mistral's flagship model with published LCA
One query, in context
Energy, carbon and water for a single use. 400-token response
About 2.2× a minute of laptop use and 15% of a phone charge.
About 9.5× a Google search and 54% of 10 m by car.
About 9.0× a teaspoon of water and 18% of a glass of water.
- Where it runsAzure France / EU regions
- Grid carbon intensity300 g CO₂/kWh · 45% renewable
- Compared with a Google search6x more than a search (0.3 Wh)
- Water~22 queries to fill 1 litre
Where Mistral Large 2 sits
#151 of 199 by energy per query. Heavier than 75% of all models. Lit dots are the same kind of model; click any dot to open it.
Want something lighter?
3.0× less energy than Mistral Large 2.
4.5× less energy than Mistral Large 2.
What a month of use adds up to
Pick how much you use it. Based on one quick use at a time, five days a week.
- Energy
- 194.8 Wh
- Carbon
- 123.4 g
- Water
- 4.87 L
A month of 5 chats a day, 5 days a week. The energy is about 1.8× a kettle boil and 3% of a UK home's day.
Work out your own →- Energy
- 779.4 Wh
- Carbon
- 493.6 g
- Water
- 19.48 L
A month of 20 chats a day, 5 days a week. The energy is about 7.1× a kettle boil and 11% of a UK home's day.
Work out your own →- Energy
- 3.90 kWh
- Carbon
- 2.47 kg
- Water
- 97.42 L
A month of 100 chats a day, 5 days a week. The energy is about 35× a kettle boil and 53% of a UK home's day.
Work out your own →How does Mistral Large 2 compare?
Next to similar models. Bars use a log scale, so each step is a big jump.
Detailed breakdown
Energy
Mistral Large 2 consumes approximately 1.8 Wh per 400-token response, based on Mistral's official Life Cycle Assessment (conducted with Carbone 4 and the French environment agency ADEME) — one of the few AI companies to publish detailed environmental impact data. Crucially, its headline 1.14 g CO2e and 45 mL figures are lifecycle (training-amortised) values: they include an allocated share of model training and hardware manufacturing, not just the direct grid emissions of a single inference. At 123B parameters, it is a dense model without MoE optimisation.
Power source & carbon
Runs primarily on Azure infrastructure in France and EU regions. France's electricity grid has a carbon intensity of approximately 50 g CO2/kWh — among the lowest in Europe — due to its heavy reliance on nuclear power (~70%). This significantly reduces the carbon footprint per query compared to models running on US grids. Note that Mistral's published 1.14 g CO2e figure is a lifecycle (training-amortised) number, so it is not directly comparable to the direct-grid CO2 figures we compute for Mistral's other models (energy × 50 gCO2/kWh).
Water use
At approximately 45 mL per 400-token response, Mistral Large 2's water figure is notably high precisely because it is a full lifecycle (cradle-to-grave) number from Mistral's Carbone 4 / ADEME LCA — it counts upstream supply-chain and manufacturing water, not just on-site cooling. It should not be compared directly with the direct-cooling water estimates we give for other models.
About Mistral Large 2
Mistral Large 2 is a 123B-parameter text and chat model from Mistral AI, released July 24, 2024. Mistral's flagship model with published LCA. At 1.8 Wh per query, it uses 6x the energy of a Google search. It runs on a Dense Transformer (decoder-only) architecture.
Mistral Large 2 in context
Key insights
- Open-source weights — can be self-hosted on infrastructure you control
- Provider-verified energy data — the highest confidence tier in our dataset
The Mistral AI Mistral family
One chart per tier, on a shared time axis. Hover a dot for details; click to open that version.
Newer in this tier: Mistral Large 4 uses 1.4 Wh per query, 22% less than Mistral Large 2.
6 versions across 2 tiers. Lightest: Mistral 7B (0.06 Wh); heaviest: Mistral Large 2 (1.8 Wh). Charts share a time axis but each tier has its own energy scale, so its trend is clear; the percentage is the change from a tier’s first version to its latest.
What does your Mistral Large 2 usage cost the planet?
Use the calculator to estimate your personal footprint based on how often you use Mistral Large 2.
Frequently asked questions
How much energy does Mistral Large 2 use per query?
Each Mistral Large 2 query consumes approximately 1.8 Wh of energy. This is 6x more than a traditional Google search (~0.3 Wh).
What is Mistral Large 2's carbon footprint?
Based on the carbon intensity of Azure France / EU regions, each query produces approximately 1.1 g of CO2. The grid in this region has a carbon intensity of 300 g CO2/kWh with 45% renewable energy.
How much water does Mistral Large 2 use?
Each query consumes approximately 45 mL of water, primarily used for cooling the data centers that process the request.
How does Mistral Large 2 compare to a Google search?
A Mistral Large 2 query uses 6x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Mistral Large 2 uses 1.8 Wh.
Technical details
- Architecture
- Dense Transformer (decoder-only)
- Parameters
- 123B
- Context window
- 128,000 tokens
- Release date
- 2024-07-24
- Open source
- Yes