OWE-Bench - Chess Puzzle Post-Training: leaderboard

OWE-Bench - Chess Puzzle Post-Training — Expert-built long-horizon science, software, ML and engineering tasks with task-specific hardware and4–18hour budgets; native baseline/ceiling normalized held-out reward,mean of8attempts and equal task means. Web frozen October1,2026;deterministic sealed verifier. Native model/effort and every attempt retained. Score fraction0–1;higher is better.

Metric: Mean normalized held-out task score (fraction, 0–1). Source: openworldengineer.mercor.com. Saturation forecast: Around February 2028. 14 models tracked. Updated 15 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Fable 5.2Max0.5171#1
2Claude Sonnet 5.5Max0.4931#19
3Claude Opus 5.5Max0.3858#5
4GPT-6.1 SolMax0.3267#10
5GPT-6 AstraMax0.3189#6
6Gemini 4 ArgonHigh0.3065#12
7Claude Fable 5.1Max0.2614#15
8Gemini 4 CarbonHigh0.2479#2
9GLM-5.3Max0.1906#68
10DeepSeek V4.1 FlashMax0.1844#62
11Muse WonderMax0.1472#28
12Beam 1.0Max0.08004#206
13Grok 4.7xHigh0.04703#47
14InklingMax0.02041#281

Interactive version: theaggregate.ai/benchmark?slug=owe-bench-chess-puzzle-post-training · How It Works · Data refreshed daily, snapshot 2026-11-16.