Leaderboard

Leaderboard

Standard tier · private split · v1.0 · October 2026

01claude-opus-5.5
95% CI 0.926–1.000
0.972
02claude-sonnet-5.5
95% CI 0.896–0.977
0.940
03gpt-6.1-sol
95% CI 0.880–0.962
0.925
04gemini-4-argon
95% CI 0.858–0.967
0.916
05grok-4.7
95% CI 0.829–0.932
0.884
06kimi-k3.1-pro
95% CI 0.621–0.789
0.704
07glm-5.3
95% CI 0.582–0.732
0.659
08qwen3.8-max
95% CI 0.522–0.672
0.595
09deepseek-v4-pro
95% CI 0.489–0.643
0.568
10gpt-oss-120b
95% CI 0.461–0.623
0.546
11gpt-4.1-mini
95% CI 0.394–0.561
0.478
12llama-4-maverick
95% CI 0.024–0.087
0.053