Cost Efficiency
Total cost per full benchmark run (lower is better)
01
GPT-5.6 Luna (medium)
$0.08
02
Gemini 3.1 Flash Lite (minimal)
$0.10
03
GPT-5.6 Luna (high)
$0.13
04
GPT-5.6 Luna (xhigh)
$0.23
05
GPT-5.4 Nano (medium)
$0.25
06
Gemini 3 Flash (dynamic)
$0.28
07
Gemini 3.1 Flash Lite (medium)
$0.32
08
Claude Haiku 4.5
$0.43
09
GPT-5.4 Mini (medium)
$0.84
10
GPT-5 Mini (medium)
$0.92
11
GPT-5.4 Mini (high)
$2.74