Models: 11 · Tiers: 29 · Samples: 90 · Invalid: 3

Day highlights

GPT-6 Astra

8.77 · max (best)
Model page

GPT-6.1 Sol

8.53 · max (best)
Model page

Gemini 4 Argon

8.13 · high (best)
Model page

high

8.13 ± 0.74
Judge scores vary between passes

low

7.80 ± 0.34
Judge scores vary between passes

GPT-6 Sol

7.88 · max (best)
Model page

GPT-5.6 Sol

7.83 · max (best)
Model page

low

6.00 ± 0.98

medium

5.46 ± 0.29
Judge scores vary between passes

GPT-6 Luna

7.11 · max (best)
Model page

Grok 4.7

6.54 · medium (best)
Model page

Grok 4.6

6.50 · high (best)
Model page

GPT-5.6 Luna

6.50 · max (best)
Model page

Grok 4.5

6.50 · high (best)
Model page

GPT-5.6 Terra

6.47 · max (best)
Model page

max

6.47 ± 1.76
Judge scores vary between passes

medium

5.97 ± 0.56
Judge scores vary between passes
Judge scores vary between passes