Leaderboard
Leaderboard
Standard tier · private split · v1.0 · October 2026
01
claude-opus-5.50.972
95% CI 0.926–1.000
02
claude-sonnet-5.50.940
95% CI 0.896–0.977
03
gpt-6.1-sol0.925
95% CI 0.880–0.962
04
gemini-4-argon0.916
95% CI 0.858–0.967
05
grok-4.70.884
95% CI 0.829–0.932
06
kimi-k3.1-pro0.704
95% CI 0.621–0.789
07
glm-5.30.659
95% CI 0.582–0.732
08
qwen3.8-max0.595
95% CI 0.522–0.672
09
deepseek-v4-pro0.568
95% CI 0.489–0.643
10
gpt-oss-120b0.546
95% CI 0.461–0.623
11
gpt-4.1-mini0.478
95% CI 0.394–0.561
12
llama-4-maverick0.053
95% CI 0.024–0.087
0.053–0.972 · score range