GameHorizon - Offline T2 Multi-Horizon Decomposition: leaderboard

Metric: Accuracy (%). Source: gamehorizon-suite.github.io. Saturation forecast: Around February 2027. 43 models tracked. Updated 18 Nov 2026.

Top models

#ModelScoreOverall rank
1Gemini 3.8 Flash81.2#39
2Gemini 3.7 Flash80.1#38
3GPT-6 Astra79.6#7
4Seed 2.1 Turbo79.4#165
5Kimi K379.3#60
6Gemini 3.6 Flash79.1#77
7GPT-5.578.9#50
8GPT-5.6 Sol77.5#33
9Qwen 3.8 27B77.4#216
10Gemini 3.5 Flash77.4#106
11Seed 2.1 Pro76.7#142
12Qwen 3.7 Plus75.9#215
13Gemini 3.1 Pro (Preview)75.8#99
14GPT-5.6 Terra75.6#69
15Seed 2.0 Pro74.9#218

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=gamehorizon-offline-t2-multi-horizon-decomposition · How It Works · Data refreshed daily, snapshot 2026-11-18.