OWE-Bench - Function Page Layout: leaderboard

OWE-Bench - Function Page Layout — 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 March 2027. 14 models tracked. Updated 15 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-6 AstraMax0.8686#6
2GPT-6.1 SolMax0.7956#10
3Claude Fable 5.2Max0.7224#1
4Gemini 4 ArgonHigh0.5646#12
5Claude Sonnet 5.5Max0.5371#19
6Muse WonderMax0.4556#28
7Gemini 4 CarbonHigh0.4404#2
8Claude Fable 5.1Max0.4358#15
9Claude Opus 5.5Max0.4#5
10GLM-5.3Max0.2646#68
11DeepSeek V4.1 FlashMax0.2535#62
12Grok 4.7xHigh0.2365#47
13Beam 1.0Max0.1161#206
14InklingMax0.07225#281

Interactive version: theaggregate.ai/benchmark?slug=owe-bench-function-page-layout · How It Works · Data refreshed daily, snapshot 2026-11-16.