AI Decathlon - Performance Optimization: leaderboard

AI Decathlon Timed Challenge (Performance Optimization) evaluating coding agents on 60-100 minute engineering, math, spatial, and interactive design tasks.

Metric: Performance Optimization Score (0-100). Source: octathlon.wendystars.com. Saturation forecast: Around May 2027. 25 models tracked. Updated 22 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-6.2 AstraxHigh80.8#4
2Claude Sonnet 5.5xHigh67.9#24
3Claude Fable 5.2xHigh64.5#2
4Gemini 4 NeonxHigh63.8#3
5Claude Opus 5.5xHigh59.2#9
6GPT-6.1 SolxHigh57.7#15
7GPT-6.2 SolxHigh57.2#10
8GPT-6 AstraxHigh55.1#11
9Gemini 4 CarbonHigh49.4#7
10Gemini 4 HydrogenxHigh48.5#16
11GPT-6.1 TerraxHigh45#25
12MiniMax-M3.1-FlashMax42.3#119
13GPT-5.6 SolxHigh39.5#36
14Kimi K3.1 ProMax37.4#35
15DeepSeek V4.1 FlashMax28.9#69

Interactive version: theaggregate.ai/benchmark?slug=ai-decathlon-performance-optimization · How It Works · Data refreshed daily, snapshot 2026-11-22.