100-Run Benchmark Summary

Baseline vs. Cognisphere ยท A1, A2, A3, T1, T2, T3

100 runs6 questions per run600 answers per mode1,200 total answers
92,917,840
tokens saved
48.00% reduction
930
inferences saved
36.90% reduction
3h 15m 42s
model time saved
42.25% reduction
Token usage for each run
Inference count for each run
Model time for each run