DeepSeek-V3 Environmental Impact

Energy Consumption

DeepSeek-V3 consumes approximately 4.2 Wh per query. Despite using a Mixture-of-Experts architecture where only a subset of parameters are active per token, the full 671B parameter model must be loaded across multiple GPUs. Jegham et al. gave it an eco-efficiency score of 0.060 — very low due to the carbon-intensive Chinese grid infrastructure.

Power Source & Carbon

DeepSeek-V3 runs on infrastructure in mainland China, sharing facilities with DeepSeek-R1 in Hangzhou and Hainan. China's electricity grid has a carbon intensity of approximately 550 g CO2/kWh — roughly 45% higher than the US average — with about 60% of electricity generated from coal.

Water Usage

At approximately 15.5 mL per query, DeepSeek-V3 has a substantial water footprint driven by both its energy requirements and the hot, humid climate at its data center locations.

How much energy does DeepSeek-V3 use per query?

Each DeepSeek-V3 query consumes approximately 4.2 Wh of energy. This is 14x more than a traditional Google search (~0.3 Wh).

What is DeepSeek-V3's carbon footprint?

Based on the carbon intensity of China (Hangzhou / Hainan), each query produces approximately 2.3 g of CO2. The grid in this region has a carbon intensity of 550 g CO2/kWh with 30% renewable energy.

How much water does DeepSeek-V3 use?

Each query consumes approximately 16 mL of water, primarily used for cooling the data centers that process the request.

How does DeepSeek-V3 compare to a Google search?

A DeepSeek-V3 query uses 14x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while DeepSeek-V3 uses 4.2 Wh.

Architecture

Transformer Mixture-of-Experts (decoder-only)

Parameters

671B

Context window

128,000 tokens

Release date

2024-12-26

Open source

Yes