GACL - Battleship: leaderboard
Game Agent Coding League: Battleship sub-benchmark. LLMs write code agents to play Battleship against each other.
Metric: Normalized tournament points (0-100); source-selected best generated agent. Source: gameagentcodingleague.com. Saturation forecast: Around January 2027. 21 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 Mini | 85.94 | #212 |
| 2 | GPT-5.5 | 85.38 | #31 |
| 3 | GPT-5.4 | 85 | #79 |
| 4 | Kimi K2.6 | 78.47 | #104 |
| 5 | GLM-5.1 | 71.19 | #97 |
| 6 | DeepSeek V4 Pro | 70.94 | #102 |
| 7 | Claude Opus 4.7 | 64.03 | #49 |
| 8 | Gemini 3.1 Pro (Preview) | 61.63 | #58 |
| 9 | DeepSeek V4 Flash | 61.41 | #137 |
| 10 | MiMo-V2.5-Pro | 59.44 | #158 |
| 11 | Claude Sonnet 4.6 | 59.25 | #90 |
| 12 | Claude Opus 4.6 | 58.81 | #63 |
| 13 | GPT-OSS-120B | 54.97 | #341 |
| 14 | Gemini 3 Flash (Preview) | 49.5 | #82 |
| 15 | Qwen 3.6 Max Preview | 41.44 | #85 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=gacl-battleship · How It Works · Data refreshed daily, snapshot 2026-10-20.