Gemini Nano Environmental Impact
Google's on-device model for Pixel and Chrome
One query, in context
Energy, carbon and water for a single use. Short on-device query (~200 tokens)
5% of a Google search.
3% of a Google search.
- Where it runsOn-device (Pixel, Samsung, Chrome)
- Grid carbon intensity458 g CO₂/kWh · 32% renewable
- Compared with a Google searchabout the same as a search
Where Gemini Nano sits
#1 of 199 by energy per query. Lighter than 100% of all models. Lit dots are the same kind of model; click any dot to open it.
Want something lighter?
Gemini Nano is already among the lighter well-known options for this kind of work, so there isn’t a clearly lighter alternative to suggest.
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
- 1.6 Wh
- Carbon
- 0.32 g
- Water
- 0.00 mL
A month of 5 chats a day, 5 days a week. The energy is 14% of a phone charge.
Work out your own →- Energy
- 6.5 Wh
- Carbon
- 1.3 g
- Water
- 0.00 mL
A month of 20 chats a day, 5 days a week. The energy is 54% of a phone charge.
Work out your own →- Energy
- 32.5 Wh
- Carbon
- 6.5 g
- Water
- 0.00 mL
A month of 100 chats a day, 5 days a week. The energy is about 2.7× a phone charge and 30% of a kettle boil.
Work out your own →How does Gemini Nano compare?
Next to similar models. Bars use a log scale, so each step is a big jump.
Detailed breakdown
Energy
Gemini Nano runs entirely on-device, consuming approximately 0.015 Wh per query — orders of magnitude less than cloud-based models. With only 3.25 billion parameters, it runs on mobile NPUs (Neural Processing Units) in devices like the Pixel 8 Pro and Samsung Galaxy S24. On-device inference eliminates network energy overhead entirely.
Power source & carbon
Running on-device, Gemini Nano's carbon footprint depends entirely on the user's local electricity grid. In a country with clean energy (e.g. France at 50 g CO2/kWh), its emissions are negligible. No data centre infrastructure is needed.
Water use
Gemini Nano uses zero water for cooling — personal devices use passive or fan-based cooling, not water-based systems. This makes it the most water-efficient model in our dataset.
About Gemini Nano
At 0.01 Wh per query, Gemini Nano is one of the most energy-efficient AI models in existence — ranked #1 across our entire dataset of 199 models. With just 3.25B parameters, it proves that useful AI doesn't have to be expensive to run. Google released it in February 8, 2024 as a text and chat model: google's on-device model for pixel and chrome.
These figures are estimates derived from hardware specifications and API benchmarks — Google has not published official energy data for Gemini Nano. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.
Key insights
- Ranked #1 most energy-efficient model in our entire dataset
- Uses less than a third of the average energy for text and chat models
The Google Gemini family
One chart per tier, on a shared time axis. Hover a dot for details; click to open that version.
16 versions across 3 tiers. Lightest: Gemini Nano (0.01 Wh); heaviest: Gemini 1.5 Pro (0.45 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 Gemini Nano usage cost the planet?
Use the calculator to estimate your personal footprint based on how often you use Gemini Nano.
Frequently asked questions
How much energy does Gemini Nano use per query?
Each Gemini Nano query consumes approximately 0.01 Wh of energy. This is about the same as a traditional Google search (~0.3 Wh).
What is Gemini Nano's carbon footprint?
Based on the carbon intensity of On-device (Pixel, Samsung, Chrome), each query produces approximately 0.00 g of CO2. The grid in this region has a carbon intensity of 458 g CO2/kWh with 32% renewable energy.
How much water does Gemini Nano use?
Each query consumes approximately 0.00 mL of water, primarily used for cooling the data centers that process the request.
How does Gemini Nano compare to a Google search?
A Gemini Nano query uses about the same as a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Gemini Nano uses 0.01 Wh.
Technical details
- Architecture
- Distilled Transformer (on-device)
- Parameters
- 3.25B
- Context window
- 32,000 tokens
- Release date
- 2024-02-08
- Open source
- No
- Training data cutoff
- 2024-06