AI Carbon Footprint Calculator
Estimate energy and CO₂ from your token usage.
With the default inputs, ~12 kg CO₂ / month — ~30 kWh for 100,000,000 tokens. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.
Results update automatically as you type.
- • Rough estimate. Real energy depends on model size, hardware and datacenter efficiency.
- Tokens / month
- 100,000,000
- Energy
- 30 kWh
- CO₂
- 12 kg
- ≈ car miles
- 30 mi
Get a rough estimate of the energy (kWh) and carbon (kg CO₂) behind your monthly token usage, based on editable energy-per-token and grid-intensity assumptions.
How this is calculated
Energy (kWh) = tokens per month ÷ 1,000 × energy per 1k tokens (Wh) ÷ 1,000. CO₂ (kg) = kWh × grid intensity (g/kWh) ÷ 1,000. Both the energy-per-token and grid-intensity assumptions are editable, so you can plug in your own figures.
Is this a good result? What to do next
This is a rough, order-of-magnitude estimate — real energy depends on model size, hardware, batching and datacenter efficiency, all of which vary widely. Use it for relative comparison and awareness, not for formal carbon accounting.
Typical planning ranges
- Default energy
- ~0.3 Wh per 1k tokens (editable)
- Default grid
- ~400 g CO₂/kWh (editable)
- Accuracy
- order-of-magnitude only
Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.
How to improve this number
- Use smaller/efficient models for high-volume tasks.
- Choose regions/providers with cleaner grids.
- Reduce tokens (caching, shorter output) to cut energy.
Common mistakes
- Treating the estimate as precise carbon accounting.
- Ignoring that model size and hardware change energy per token drastically.
When to use a different approach
To translate usage into dollars, use the tokens-per-dollar or LLM API cost calculators. For local hardware, see the GPU calculators.
Worked example (defaults)
With the default inputs above, here is the result:
- • Rough estimate. Real energy depends on model size, hardware and datacenter efficiency.
- Tokens / month
- 100,000,000
- Energy
- 30 kWh
- CO₂
- 12 kg
- ≈ car miles
- 30 mi
Frequently asked questions
How accurate is this?+
It's a rough order-of-magnitude estimate. Real energy depends on model size, hardware, batching and datacenter efficiency.
Where do the defaults come from?+
Published estimates for inference energy per token and average grid carbon intensity. Adjust them to your own figures.