Game Agent Coding League — leaderboard
LLMs compete in 7 classic board games (Battleship, Tic-Tac-Toe, Connect4, Wizard, WordMatrix, MiniChess, Surround Morris). Overall weighted score measures game-playing ability.
Metric: Weighted Score (0-100). Source: gameagentcodingleague.com. Status: saturation imminent. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 73.21 |
| 2 | GPT-5.4 | 71.93 |
| 3 | Claude Opus 4.7 | 66.96 |
| 4 | Gemini 3.1 Pro (Preview) | 62.79 |
| 5 | Claude Opus 4.6 | 62.32 |
| 6 | GLM-5.1 | 61.39 |
| 7 | Kimi K2.6 | 61.09 |
| 8 | Claude Sonnet 4.6 | 59.71 |
| 9 | MiMo-V2.5-Pro | 58.22 |
| 10 | DeepSeek V4 Pro | 57.74 |
| 11 | Qwen 3.6 Plus | 53.16 |
| 12 | Gemini 3 Flash (Preview) | 51.55 |
| 13 | Qwen 3.6 Max Preview | 49.26 |
| 14 | GPT-5.4 Mini | 49.09 |
| 15 | DeepSeek V4 Flash | 46.02 |
Interactive version: theaggregate.ai/benchmark?slug=game-agent-coding-league · How the rankings work · Data refreshed daily, snapshot 2026-07-22.