GACL - Backgammon: leaderboard

Metric: Normalized tournament points (0-100); source-selected best generated agent. Source: gameagentcodingleague.com. Saturation forecast: Around January 2027. 21 models tracked.

Top models

#ModelScore
1GPT-5.487.5
2Claude Opus 4.785.94
3GLM-5.183.75
4GPT-5.579.69
5Claude Opus 4.675
6Gemini 3.1 Pro (Preview)74.38
7GPT-5.4 Mini73.75
8Claude Sonnet 4.673.44
9Qwen 3.6 Plus70.63
10MiMo-V2.5-Pro70.63
11Grok 4.2066.88
12Gemini 3 Flash (Preview)61.56
13Kimi K2.656.88
14Qwen 3.6 Max Preview53.75
15DeepSeek V4 Pro46.88

Interactive version: theaggregate.ai/benchmark?slug=gacl-backgammon · How It Works · Data refreshed daily, snapshot 2026-10-09.