2.8× less energy than Qwen3.8-Max.
Qwen3.8-Max Environmental Impact
Alibaba's largest model — 2.4T-parameter MoE with 95B active
The figures below are for one quick question. A longer chat with follow-ups uses about 1.9 Wh, and an hour or two of agent work about 55.0 Wh.
One query, in context
Energy, carbon and water for a single use. Quick single question (one exchange)
About 1.8× a Google search and 66% of a minute of laptop use.
About 2.4× a Google search and 14% of 10 m by car.
About 40× a drop of water and 40% of a teaspoon of water.
- Where it runsAlibaba Cloud (China) / self-hosted
- Grid carbon intensity525 g CO₂/kWh · 33% renewable
- Compared with a Google search2x more than a search (0.3 Wh)
- Water~500 queries to fill 1 litre
Where Qwen3.8-Max sits
#122 of 199 by energy per query. Heavier than 61% of all models. Lit dots are the same kind of model; click any dot to open it.
Want something lighter?
2.3× less energy than Qwen3.8-Max.
7.9× less energy than Qwen3.8-Max.
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
- 59.5 Wh
- Carbon
- 31.4 g
- Water
- 217 mL
A month of 5 chats a day, 5 days a week. The energy is about 5.0× a phone charge and 54% of a kettle boil.
Work out your own →- Energy
- 238.2 Wh
- Carbon
- 125.6 g
- Water
- 866 mL
A month of 20 chats a day, 5 days a week. The energy is about 2.2× a kettle boil and 3% of a UK home's day.
Work out your own →- Energy
- 1.19 kWh
- Carbon
- 627.8 g
- Water
- 4.33 L
A month of 100 chats a day, 5 days a week. The energy is about 11× a kettle boil and 16% of a UK home's day.
Work out your own →How does Qwen3.8-Max compare?
Next to similar models. Bars use a log scale, so each step is a big jump.
Detailed breakdown
Energy
We estimate Qwen3.8-Max at ~0.55 Wh for a quick single question. It has not been benchmarked directly; the closest measured model is Kimi K3 (2.8T parameters), which CLEER (Sustainable AI Group, Sept 2026) ran on B300 accelerators at ~0.65 Wh. Qwen3.8-Max is somewhat smaller at 2.4T total and ~95B active parameters, so we scale down modestly. For comparison CLEER measured the much smaller Qwen3.5-397B at ~0.13 Wh — a reminder that the jump to trillion-parameter open models carries a real energy cost.
Power source & carbon
Hosted inference runs on the Chinese grid at ~525 gCO2/kWh (Ember 2026). Open weights mean it is also widely self-hosted, where the footprint depends entirely on the local grid.
Water use
At ~2 mL per quick question on Chinese infrastructure (3.7 L/kWh combined water factor).
About Qwen3.8-Max
Qwen3.8-Max is a 2400B-parameter text and chat model from Alibaba, released August 3, 2026. Alibaba's largest model — 2.4T-parameter MoE with 95B active. At 0.55 Wh per query, it uses roughly the same energy as a Google search. It runs on a Mixture-of-Experts (2.4T total, ~95B active) architecture.
These figures are estimates derived from hardware specifications and API benchmarks — Alibaba has not published official energy data for Qwen3.8-Max. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.
Qwen3.8-Max in context
Key insights
- Open-source weights — can be self-hosted on infrastructure you control
What does your Qwen3.8-Max usage cost the planet?
Use the calculator to estimate your personal footprint based on how often you use Qwen3.8-Max.
Frequently asked questions
How much energy does Qwen3.8-Max use per query?
Each Qwen3.8-Max query consumes approximately 0.55 Wh of energy. This is 2x more than a traditional Google search (~0.3 Wh).
What is Qwen3.8-Max's carbon footprint?
Based on the carbon intensity of Alibaba Cloud (China) / self-hosted, each query produces approximately 0.29 g of CO2. The grid in this region has a carbon intensity of 525 g CO2/kWh with 33% renewable energy.
How much water does Qwen3.8-Max use?
Each query consumes approximately 2 mL of water, primarily used for cooling the data centers that process the request.
How does Qwen3.8-Max compare to a Google search?
A Qwen3.8-Max query uses 2x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Qwen3.8-Max uses 0.55 Wh.
Technical details
- Architecture
- Mixture-of-Experts (2.4T total, ~95B active)
- Parameters
- 2400B
- Release date
- 2026-08-03
- Open source
- Yes