Mixtral 8x22B Environmental Impact
Large MoE model rivalling GPT-4 class performance
- Architecture
- Sparse Mixture-of-Experts (8 experts, 2 active)
- Parameters
- 176B
- Context
- 65,000 tokens
- Provider
- Mistral AI
Energy per query
1.2 Wh
4x more than a Google search (0.3 Wh)
CO2 per query
0.76 g
Global Average grid (475 gCO₂/kWh)
Water per query
5 mL
~200 queries to fill 1 litre
Processing location
Self-hosted (varies, requires multi-GPU)
Provider
Mistral AI
Category
Text / Chat
Grid carbon intensity
475 g CO2/kWh (27% renewable)
How does Mixtral 8x22B compare?
Detailed Breakdown
Energy Consumption
Mixtral 8x22B has 176B total parameters but only activates ~44B per token. At ~1.2 Wh per query, it matches GPT-4 Turbo class performance while using MoE efficiency. Requires multi-GPU setups but significantly less compute than an equivalent dense model.
Power Source & Carbon
Open-source (Apache 2.0). Typically deployed on cloud GPU clusters due to its size requirements.
Water Usage
At ~5.0 mL per query, the water footprint reflects its multi-GPU inference requirements.
About Mixtral 8x22B
Mixtral 8x22B is an open-source text and chat model from Mistral AI, released in April 17, 2024, that runs well below the category average for energy consumption at 1.2 Wh per query. Because its weights are publicly available, it can be self-hosted on any infrastructure — meaning its carbon footprint depends entirely on where and how you choose to run it. At 176B parameters, it large moe model rivalling gpt-4 class performance.
These figures are estimates derived from hardware specifications and API benchmarks — Mistral AI has not published official energy data for Mixtral 8x22B. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.
Key Insights
What does your Mixtral 8x22B usage cost the planet?
Use our calculator to estimate your personal environmental footprint based on how often you use Mixtral 8x22B.
Calculate My ComputeFrequently Asked Questions
How much energy does Mixtral 8x22B use per query?
Each Mixtral 8x22B query consumes approximately 1.2 Wh of energy. This is 4x more than a traditional Google search (~0.3 Wh).
What is Mixtral 8x22B's carbon footprint?
Based on the carbon intensity of Self-hosted (varies, requires multi-GPU), each query produces approximately 0.76 g of CO2. The grid in this region has a carbon intensity of 475 g CO2/kWh with 27% renewable energy.
How much water does Mixtral 8x22B use?
Each query consumes approximately 5 mL of water, primarily used for cooling the data centers that process the request.
How does Mixtral 8x22B compare to a Google search?
A Mixtral 8x22B query uses 4x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Mixtral 8x22B uses 1.2 Wh.
Technical Details
Architecture
Sparse Mixture-of-Experts (8 experts, 2 active)
Parameters
176B
Context window
65,000 tokens
Release date
2024-04-17
Open source
Yes
Training data cutoff
2024-04