Droplet methodology
Model version 0.1.0 · Last updated: August 8, 2026
How the estimate is calculated
Droplet estimates output tokens from the character length of each completed assistant response. It then estimates the energy for that response and converts the energy estimate into on-site and electricity-generation water estimates.
tokens_out = char_count / 4 E_Wh = (Wh_base + tokens_out × Wh_per_token) × reasoning_multiplier water_onsite = E_Wh × WUE_dc water_grid = E_Wh × EWIF_grid water_total = water_onsite + water_grid
On-site and grid water
On-site water is used directly for data-center cooling. Grid water is used upstream to generate the electricity that powers a request. “On-site cooling only” shows the direct estimate; “including electricity generation” combines both values.
Coefficients
| Coefficient | Low | Mid | High | Unit |
|---|---|---|---|---|
| Wh_base | 0.02 | 0.05 | 0.30 | Wh / request |
| Wh_per_token | 0.0002 | 0.00063 | 0.009 | Wh / token |
| reasoning_multiplier | 1.5 | 3 | 10 | × |
| WUE_dc | 0.2 | 1.1 | 1.9 | mL / Wh |
| EWIF_grid | 1.0 | 4.5 | 6.5 | mL / Wh |
The low, mid, and high ranges reflect uncertainty across hardware efficiency, response length, data-center water use, and electricity-generation water use.
Research anchors
Our low estimate follows Google’s production measurements of on-site cooling water. Our high estimate follows Li et al., which adds the water consumed generating the electricity. The truth for any given prompt sits somewhere between, and depends on which data center served you—something neither we nor you can know.
-
Elsworth, C., Huang, K., Patterson, D., Schneider, I., Sedivy, R., Goodman, S., Townsend, B., Ranganathan, P., Dean, J., Vahdat, A., Gomes, B., & Manyika, J. (2025). Measuring the environmental impact of delivering AI at Google Scale. arXiv:2508.15734.
-
Li, P., Yang, J., Islam, M. A., & Ren, S. (2025). Making AI Less “Thirsty”: Uncovering and Addressing the Secret Water Footprint of AI Models. Communications of the ACM, 68(7), 54–63.
Communications of the ACM (peer-reviewed version) · arXiv preprint
Known limitations
- Reasoning tokens are not visible, so Droplet applies a multiplier only when reasoning UI is detected.
- Droplet does not know the data-center region that served a request; water and grid intensity vary by region.
- Token count is approximated from characters, so it is an estimate rather than a measured token total.
- Water embodied in chip manufacturing and data-center construction is not included.
- Input-token processing is included in the base estimate rather than separately measured.
- Host-side retrieval, image generation, and tool calls are not included.