GACL - Connect4 — leaderboard
Game Agent Coding League: Connect4 sub-benchmark. LLMs write code agents to play Connect Four against each other.
Metric: Normalized Score (0-100). Source: gameagentcodingleague.com. Status: saturation imminent. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 76.3 |
| 2 | Claude Opus 4.7 | 73.23 |
| 3 | GLM-5.1 | 71.2 |
| 4 | GPT-5.5 | 69.64 |
| 5 | GPT-5.4 | 69.32 |
| 6 | Kimi K2.6 | 66.46 |
| 7 | Claude Opus 4.6 | 65.57 |
| 8 | Claude Sonnet 4.6 | 57.81 |
| 9 | Qwen 3.6 Max Preview | 56.41 |
| 10 | Qwen 3.6 35B A3B | 55.83 |
| 11 | Qwen 3.6 Plus | 55 |
| 12 | DeepSeek V4 Pro | 54.37 |
| 13 | MiMo-V2.5-Pro | 45.05 |
| 14 | Qwen 3.6 27B | 43.54 |
| 15 | MiniMax-M2.7 | 42.76 |
Interactive version: theaggregate.ai/benchmark?slug=gacl-connect4 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.