Veo 2 Environmental Impact

Energy Consumption

Veo 2 consumes approximately 350 Wh per ~8-second video clip. While comparable in scale to OpenAI's Sora, Veo 2 benefits from Google's TPU infrastructure which is generally more power-efficient than GPU-based inference. Video generation remains one of the most energy-intensive AI tasks.

Power Source & Carbon

Runs on Google's global TPU infrastructure with 100% renewable energy matching and a PUE of 1.10. Google's lower carbon intensity (130 g CO2/kWh market-based) significantly reduces the per-query emissions compared to Sora on Azure.

Water Usage

At approximately 380 mL per video clip, Veo 2's water consumption is substantial but significantly lower than Sora's estimate, partly due to Google's more efficient cooling infrastructure.

How much energy does Veo 2 use per query?

Each Veo 2 query consumes approximately 350.0 Wh of energy. This is 1167x more than a traditional Google search (~0.3 Wh).

What is Veo 2's carbon footprint?

Based on the carbon intensity of Google global network (64% renewable), each query produces approximately 46.0 g of CO2. The grid in this region has a carbon intensity of 300 g CO2/kWh with 64% renewable energy.

How much water does Veo 2 use?

Each query consumes approximately 380 mL of water, primarily used for cooling the data centers that process the request.

How does Veo 2 compare to a Google search?

A Veo 2 query uses 1167x more than a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Veo 2 uses 350.0 Wh.

Architecture

Diffusion Transformer (video)

Release date

2024-12-16

Open source

No