AI energy and carbon calculator: kWh and CO2e per million tokens
Estimate operational electricity and carbon from GPU power, throughput, utilization, PUE and grid intensity. The defaults are planning assumptions; replace them with your own measurements.
Workload and facility
Advanced: batch mode and custom carbon intensity
Batch mode only changes the utilization assumption. It does not assume a lower carbon grid. If you schedule batch work into a cleaner hour, enter that hour's intensity in the custom field.
Throughput is an input, not a benchmark. Use a number you measured on the same GPU, model, quantization and context length. The default 1,000 tokens/sec/GPU is a planning value, not a claim about any model.
Result real-time
Operational electricity only: GPU TDP × utilization × PUE. It excludes embodied carbon, networking, storage, water and training. Arithmetic, not a measurement.
Energy per million tokens is a division problem
The core formula is small enough to check by hand:
facility_power_W = gpu_tdp_W × gpu_count × utilization × PUE
kWh_per_1M_tokens = facility_power_W ÷ (tokens_per_second × 3.6)
The 3.6 is unit conversion: one watt for one hour is 3.6 kJ, and one kWh is 3.6 MJ. With one 700 W GPU at 45% utilization and PUE 1.52, facility power is 479 W. At 1,000 tokens per second, that is 479 ÷ 3,600 = 0.133 kWh per million tokens. Double the throughput and the energy per token halves; double the PUE and it doubles.
This is why “kWh per 1M tokens” is not a property of a model. It is a property of the serving stack at a moment in time. A 70B model on an idle H100 and a 7B model on a saturated L4 can have the same energy per token if their throughput-to-power ratios match. The model matters because it changes throughput and memory pressure, not because it has an intrinsic energy label.
Thirty GPUs at four throughput levels
The table below applies the formula to each GPU's published TDP at 45% utilization and PUE 1.52. The throughput columns are assumptions. They let you compare the shape of the trade-off, not claim that a GPU reaches any particular speed on your model.
| GPU | Memory GB | Bandwidth GB/s | TDP W | 250 tok/s | 500 tok/s | 1,000 tok/s | 2,000 tok/s |
|---|---|---|---|---|---|---|---|
| NVIDIA H100 SXM | 80 | 3,350 | 700 | 0.532 | 0.266 | 0.133 | 0.067 |
| NVIDIA H100 PCIe | 80 | 2,000 | 350 | 0.266 | 0.133 | 0.067 | 0.033 |
| NVIDIA H200 SXM | 141 | 4,800 | 700 | 0.532 | 0.266 | 0.133 | 0.067 |
| NVIDIA B200 SXM | 192 | 8,000 | 1000 | 0.760 | 0.380 | 0.190 | 0.095 |
| NVIDIA A100 SXM 80GB | 80 | 2,039 | 400 | 0.304 | 0.152 | 0.076 | 0.038 |
| NVIDIA A100 PCIe 80GB | 80 | 1,935 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA A100 SXM 40GB | 40 | 1,555 | 400 | 0.304 | 0.152 | 0.076 | 0.038 |
| NVIDIA A30 | 24 | 933 | 165 | 0.125 | 0.063 | 0.031 | 0.016 |
| NVIDIA A10 | 24 | 600 | 150 | 0.114 | 0.057 | 0.028 | 0.014 |
| NVIDIA A10G | 24 | 600 | 150 | 0.114 | 0.057 | 0.028 | 0.014 |
| NVIDIA L40 | 48 | 864 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA L40S | 48 | 864 | 350 | 0.266 | 0.133 | 0.067 | 0.033 |
| NVIDIA L4 | 24 | 300 | 72 | 0.055 | 0.027 | 0.014 | 0.007 |
| NVIDIA T4 | 16 | 320 | 70 | 0.053 | 0.027 | 0.013 | 0.007 |
| NVIDIA V100 SXM2 32GB | 32 | 900 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA V100 PCIe 32GB | 32 | 900 | 250 | 0.190 | 0.095 | 0.048 | 0.024 |
| NVIDIA RTX 4090 | 24 | 1,008 | 450 | 0.342 | 0.171 | 0.086 | 0.043 |
| NVIDIA RTX 4080 Super | 16 | 736 | 320 | 0.243 | 0.122 | 0.061 | 0.030 |
| NVIDIA RTX 3090 | 24 | 936 | 350 | 0.266 | 0.133 | 0.067 | 0.033 |
| NVIDIA RTX 3080 | 10 | 760 | 320 | 0.243 | 0.122 | 0.061 | 0.030 |
| NVIDIA RTX 5090 | 32 | 1,792 | 575 | 0.437 | 0.218 | 0.109 | 0.055 |
| NVIDIA RTX 5080 | 16 | 960 | 360 | 0.274 | 0.137 | 0.068 | 0.034 |
| NVIDIA RTX 5070 Ti | 16 | 896 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA RTX A6000 | 48 | 768 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA RTX 6000 Ada | 48 | 960 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| NVIDIA RTX 5000 Ada | 32 | 576 | 250 | 0.190 | 0.095 | 0.048 | 0.024 |
| AMD Instinct MI300X | 192 | 5,300 | 750 | 0.570 | 0.285 | 0.142 | 0.071 |
| AMD Instinct MI250X | 128 | 3,200 | 560 | 0.426 | 0.213 | 0.106 | 0.053 |
| AMD Instinct MI210 | 64 | 1,600 | 300 | 0.228 | 0.114 | 0.057 | 0.028 |
| Intel Gaudi 3 | 128 | 3,700 | 600 | 0.456 | 0.228 | 0.114 | 0.057 |
At 250 tokens/sec/GPU, a 700 W H100 uses about 0.53 kWh per million tokens; at 2,000 tokens/sec/GPU it uses about 0.07 kWh. The 7.6× spread is throughput, not hardware efficiency.
Model size changes the GPU count, not the formula
A model has to fit in GPU memory before it can serve a token. At 4-bit quantization, weights take roughly 0.5 GB per billion parameters, plus KV cache and runtime overhead. That gives a minimum GPU count, and the count multiplies facility power and throughput together. The table uses an 80 GB GPU and a 90% memory budget; it is a sizing check, not a throughput claim.
| Model | 4-bit weights | Minimum GPUs | GPUs | Facility kW | M tokens/hour at 1,000 tok/s/GPU | kWh / 1M tokens |
|---|---|---|---|---|---|---|
| 7B | ≈4 GB | 1 × 24 GB | 1 | 0.479 | 3.6 | 0.133 |
| 13B | ≈7 GB | 1 × 24 GB | 1 | 0.479 | 3.6 | 0.133 |
| 32B | ≈18 GB | 1 × 24 GB | 1 | 0.479 | 3.6 | 0.133 |
| 70B | ≈39 GB | 1 × 80 GB | 1 | 0.479 | 3.6 | 0.133 |
| 120B | ≈66 GB | 1 × 80 GB | 1 | 0.479 | 3.6 | 0.133 |
| 405B | ≈223 GB | 3 × 80 GB | 3 | 1.436 | 10.8 | 0.133 |
Energy per token stays flat because power and throughput both scale with GPU count. The 405B row draws three times the power and serves three times the tokens at the same per-GPU rate. Real tensor-parallel serving adds interconnect power and rarely scales perfectly; use a measurement when you have one.
PUE: the facility multiplier
PUE is total facility power divided by IT power. A PUE of 1.52 means every 1 W of GPU power pulls 1.52 W from the grid, with the extra going to cooling, power distribution and losses. The Uptime Institute's 2026 survey reports an industry-wide annual average of 1.52; facilities commissioned in the previous five years average 1.48, and large facilities of 20 MW or more average 1.44.
Use the number your facility publishes, and check whether it is annual, monthly or instantaneous. A design PUE is not an operating PUE. For a colocation quote, ask for the contracted PUE and the measurement boundary. For a cloud GPU, the provider may publish a region-level PUE; if it does not, 1.5 is a defensible planning default, not a fact about that provider.
Grid carbon intensity: annual averages, not hourly truth
The carbon table uses Ember's latest annual intensity values via Our World in Data. The spread is large: Sweden and Norway are under 40 gCO2e/kWh, France is about 41, the United States about 384, India about 670. At 0.133 kWh per million tokens, the same workload emits roughly 5 g in Sweden and 89 g in India.
| Region | gCO2e/kWh | gCO2e / 1M tokens | kgCO2e / 1B tokens |
|---|---|---|---|
| Sweden | 35 | 5 | 0.005 |
| Norway | 28 | 4 | 0.004 |
| France | 41 | 6 | 0.006 |
| Brazil | 110 | 15 | 0.015 |
| Spain | 154 | 20 | 0.020 |
| Canada | 191 | 25 | 0.025 |
| United Kingdom | 217 | 29 | 0.029 |
| Netherlands | 254 | 34 | 0.034 |
| Ireland | 257 | 34 | 0.034 |
| Italy | 285 | 38 | 0.038 |
| Germany | 330 | 44 | 0.044 |
| United States | 384 | 51 | 0.051 |
| South Korea | 417 | 55 | 0.055 |
| Japan | 477 | 63 | 0.063 |
| Singapore | 497 | 66 | 0.066 |
| Australia | 525 | 70 | 0.070 |
| China | 525 | 70 | 0.070 |
| Poland | 589 | 78 | 0.078 |
| India | 670 | 89 | 0.089 |
| World | 458 | 61 | 0.061 |
Source: Ember via Our World in Data, “Carbon intensity of electricity”, retrieved 20 November 2026. Values are national annual averages; subnational grids and hourly marginal intensity can differ substantially.
Batch versus real-time
Real-time chat keeps a GPU resident and often runs at low utilization between requests. Batch work can queue enough requests to keep the same GPU busy, which raises utilization and improves tokens per kWh, provided the GPU does not hit a power or thermal limit. The calculator's batch mode changes utilization from 45% to 85%; it does not change the carbon intensity unless you enter a different value.
There are two separate savings. The first is utilization: the fixed power of an idle GPU is spread over more tokens. The second is carbon-aware scheduling: running deferrable work when the grid is cleaner. The second only helps if the workload can wait and the grid signal is real. A batch job that must finish in five minutes cannot wait for a windy hour.
Do not double-count. If you raise utilization from 45% to 85%, facility power rises by 89%, but throughput may rise by more than 89% because the GPU was previously idle. Measure both before and after; the ratio is the only honest efficiency claim.
Methodology and limits
The calculator uses published GPU TDP, a user-supplied throughput, a user-supplied utilization and PUE, and a regional annual carbon intensity. It does not model GPU frequency/voltage curves, memory power, CPU host power, networking, storage or cooling water. It does not include embodied emissions from manufacturing. It does not assume that any GPU achieves a particular tokens-per-second figure on any model.
For a procurement estimate, replace the default throughput with a measurement from your own stack at your p95 context length. Then replace the default PUE with the facility's reported value. The remaining uncertainty is the grid; use the annual average for reporting and an hourly signal for scheduling. Related: renting GPU capacity, local AI deployment, and what actually drives inference cost per million tokens.
FAQ
How many kWh does 1 million tokens use?
It depends on throughput and facility overhead, not on the model name. At 700 W per GPU, 45% utilization and a PUE of 1.52, a GPU drawing 479 W and serving 1,000 tokens per second uses about 0.13 kWh per million tokens. At 250 tokens per second the same GPU uses about 0.53 kWh. Enter your own throughput; the calculator does not assume a benchmark.
What PUE should I use?
Use the PUE your facility reports. The Uptime Institute's 2026 survey puts the industry-wide annual average at 1.52, with newer and larger facilities often between 1.3 and 1.48. A hyperscale facility may be lower; a small on-premises server room may be well above 1.5. PUE multiplies the GPU power, so a 0.2 difference changes the result by about 13%.
Does batching reduce energy per token?
It can, but not automatically. Batching raises GPU utilization and spreads the fixed cost of weights and idle power across more tokens. If the GPU is already saturated, adding batch does not reduce the energy per token; it increases throughput at roughly the same power. The saving comes from running the same hardware closer to its power-performance sweet spot, not from the word “batch”.
Where does the carbon intensity data come from?
The regional values are Ember's latest annual carbon-intensity figures, published through Our World in Data and retrieved on 20 November 2026. They are annual averages for 2025, not hourly marginal emissions. A grid can be much cleaner or dirtier at a particular hour, so use the custom-intensity field if you are scheduling batch work.
Does this include embodied carbon or water use?
No. The calculator covers operational electricity for the GPUs plus the facility overhead captured by PUE. It does not include GPU manufacturing, server networking, storage, cooling water, or the energy used to train the model. Those are real costs, but they need a different dataset and a different boundary.
Sources and verification
Primary sources were opened and checked on 20 November 2026. Product terms and datasets change; the linked source is authoritative.
- Uptime Institute Global Data Center Survey 2026 — industry average PUE and the larger-facility figure checked 20 November 2026
- Ember via Our World in Data: carbon intensity of electricity — annual national carbon-intensity values, latest year 2025 checked 20 November 2026
- NVIDIA H100 datasheet — H100 SXM and PCIe memory, bandwidth and TDP checked 20 November 2026
- AMD Instinct MI300X specifications — MI300X memory, bandwidth and board power checked 20 November 2026
Skip the hardware and the PUE spreadsheet
Plugsky serves hosted models through one OpenAI-compatible API. You get published plan caps instead of a GPU bill, and the free tier starts at 100 requests and 50K tokens per day.
See plans → Browse models