Greenpixie

AI Explorer

Energy, carbon and water per 1,000 tokens for every model the Greenpixie AI Explorer covers. Energy is split three ways: the hardware serving the model, the data-centre overhead on top of it, and their total.

Model

Providers

Loading…

Batch size

How densely requests are batched on the serving hardware — higher batching spreads the energy across more tokens.

Code

curl -G "https://api.greenpixie.com/v1/ratecards/ai" \
  --data-urlencode "effective_batch_size=medium" \
  --data-urlencode "limit=1000" \
  -H "Authorization: Bearer $GPX_KEY"

This request returns every model — the provider and model-search filters narrow the results in your browser, not in the API call.

Loading the AI Explorer…