3.9× less energy than Phi-4 Reasoning.
Phi-4 Reasoning Environmental Impact
14B reasoning model that outperforms o1-mini and DeepSeek-R1-Distill-70B
The figures below are for one quick question. A longer chat with follow-ups uses about 0.95 Wh, and an hour or two of agent work about 27.0 Wh.
One query, in context
Energy, carbon and water for a single use. Quick single question (one exchange)
90% of a Google search.
About 1.0× a Google search and 6% of 10 m by car.
About 19× a drop of water and 19% of a teaspoon of water.
- Where it runsSelf-hosted / Azure
- Grid carbon intensity458 g CO₂/kWh · 32% renewable
- Compared with a Google searchabout the same as a search (0.3 Wh)
- Water~1,031 queries to fill 1 litre
Where Phi-4 Reasoning sits
#74 of 199 by energy per query. Lighter than 63% of all models. Lit dots are the same kind of model; click any dot to open it.
Want something lighter?
1.8× less energy than Phi-4 Reasoning.
4.5× less energy than Phi-4 Reasoning.
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
- 29.2 Wh
- Carbon
- 13.0 g
- Water
- 105 mL
A month of 5 chats a day, 5 days a week. The energy is about 2.4× a phone charge and 27% of a kettle boil.
Work out your own →- Energy
- 116.9 Wh
- Carbon
- 52.0 g
- Water
- 420 mL
A month of 20 chats a day, 5 days a week. The energy is about 1.1× a kettle boil and 2% of a UK home's day.
Work out your own →- Energy
- 584.6 Wh
- Carbon
- 259.8 g
- Water
- 2.10 L
A month of 100 chats a day, 5 days a week. The energy is about 5.3× a kettle boil and 8% of a UK home's day.
Work out your own →How does Phi-4 Reasoning compare?
Next to similar models. Bars use a log scale, so each step is a big jump.
Detailed breakdown
Energy
We estimate Phi-4 Reasoning at ~0.27 Wh for a quick single question. Phi-4 Reasoning is the 14B Phi-4 with always-on thinking. Standard Phi-4 comes to ~0.09 Wh on the pattern across the open models CLEER has benchmarked (roughly 0.02 Wh plus 0.005 Wh per billion active parameters); we multiply by 3 for the extra reasoning tokens a small model generates on a quick question. This is our extrapolation, not a CLEER figure or a measurement. It replaces our earlier ~1.8 Wh estimate, which was built on 2025 benchmark figures for older hardware.
Power source & carbon
Open-source. Can be self-hosted on a single GPU, making it one of the most accessible reasoning models.
Water use
At ~0.97 mL per quick question, derived from the energy estimate at the ~3.6 L/kWh water factor we apply to this model's infrastructure. Water scales directly with energy, so longer chats, reasoning and agentic sessions use proportionally more.
About Phi-4 Reasoning
Phi-4 Reasoning is an open-source text and chat model from Microsoft, released in May 10, 2025, that runs below the category average for energy consumption at 0.27 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 14B parameters, it 14b reasoning model that outperforms o1-mini and deepseek-r1-distill-70b.
These figures are estimates derived from hardware specifications and API benchmarks — Microsoft has not published official energy data for Phi-4 Reasoning. Actual consumption may vary significantly depending on batching, quantisation, and infrastructure optimisations that we cannot observe from outside.
Phi-4 Reasoning in context
Key insights
- Open-source weights — can be self-hosted on infrastructure you control
The Microsoft Phi family
One chart per tier, on a shared time axis. Hover a dot for details; click to open that version.
What does your Phi-4 Reasoning usage cost the planet?
Use the calculator to estimate your personal footprint based on how often you use Phi-4 Reasoning.
Frequently asked questions
How much energy does Phi-4 Reasoning use per query?
Each Phi-4 Reasoning query consumes approximately 0.27 Wh of energy. This is about the same as a traditional Google search (~0.3 Wh).
What is Phi-4 Reasoning's carbon footprint?
Based on the carbon intensity of Self-hosted / Azure, each query produces approximately 0.12 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 Phi-4 Reasoning use?
Each query consumes approximately 0.97 mL of water, primarily used for cooling the data centers that process the request.
How does Phi-4 Reasoning compare to a Google search?
A Phi-4 Reasoning query uses about the same as a Google search in terms of energy. A Google search uses approximately 0.3 Wh, while Phi-4 Reasoning uses 0.27 Wh.
Technical details
- Architecture
- Dense Transformer with chain-of-thought reasoning
- Parameters
- 14B
- Context window
- 128,000 tokens
- Release date
- 2025-05-10
- Open source
- Yes
- Training data cutoff
- 2025-04