⚑ LLM Energy Testing Lab

Measure real-world energy consumption and carbon footprint of LLM interactions

πŸ“š
Documentation

🎯 Prompt Model β“˜

πŸ’‘ This is your actual query - the only "original" tokens

🧩 Hidden Middleware β“˜

⚠️ Middleware Overhead: Each injection adds hidden tokens to your query

Inputs (Prefill Stage)

Outputs (Decode Stage)

0.7

πŸ’¬ Response Output

Run a test to see the model response here.

πŸ“Š Token Analysis β“˜

Input
-
Overhead
-
Thinking
-
Output
-
Strategy
-
Total Tokens
-
Latency
-
Tokens/Sec
-

πŸ“ˆ Tests Overview β“˜

⚑
Total Energy
-
Watt-hours per run
🎯
Intensity
-
Wh / 1k tokens
πŸš€
Speed
-
Total time (s)
🌱
Carbon Footprint
-
gCO2 emissions
Metric:
Norm:

πŸ“ Benchmark Estimation β“˜

Current Benchmark

-

Add Custom Benchmark

Uses the selected benchmark and external baselines to estimate Wh/1000 e-tokens for this run and session.

⚑ Live Power RAPL Results β“˜

πŸ”„ Batch Testing β“˜

🎲 Reproducibility & Calibration β“˜

Not run yet.