GPT-6 Astrasuspected nerf 13.41

OpenAI · Codex · 272k

Score
13.41
Rank
9/ 10
Best on
0/ 8 tasks
API cost
$1158
Input tokens · 97.4% cached
776 M
Output tokens
3.4 M
TaskJudgedRankTierSelf-bestRoundsHoursAPI cost
LLaVA1.41×9 / 10N11.29×4010.0 h$187trace →
YOCO1.79×5 / 10N11.79×2510.0 h$99.7trace →
NanoChat1.13×4 / 10N11.21×2710.0 h$143trace →
BAGEL SFT1.62×5 / 10N11.65×1510.0 h$78.9trace →
R31.47×7 / 10N01.54×2210.0 h$89.3trace →
3DGS2.04×7 / 10N12.07×2210.0 h$97.5trace →
Slime GRPO1.15×10 / 10N11.13×2010.0 h$217trace →
Slime OPD1.30×9 / 10N11.38×2210.0 h$245trace →
↑ Leaderboard