Age of LLM: leaderboard
We introduce Age of LLM, a turn-based 1v1 benchmark in which two LLMs face off on a 13x7 grid to destroy the enemy base. Three stressors are deliberate: fog of war, full diplomacy (messages, ceasefires, ultimatums; uranium kept secret), and a reliability dimension where every turn must follow a strict JSON schema and an illegal action is silently discarded. The engine is private and each match use
Metric: Points per match (self-reported). Source: benchmarklist.com. Saturation forecast: Around October 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 3 | #31 |
| 2 | Gemini 3.5 Flash | 2.4 | #60 |
| 3 | MiniMax-M3 | 2.14 | #136 |
| 4 | Claude Fable 5 | 2 | #16 |
| 5 | DeepSeek V4 Pro | 1.87 | #102 |
| 6 | GLM-5.1 | 1.8 | #97 |
| 7 | Claude Opus 4.8 | 1.67 | #37 |
| 8 | GLM-5.2 | 1.67 | #73 |
| 9 | Gemini 3.1 Pro (Preview) | 1.2 | #58 |
| 10 | Nemotron 3 Ultra | 0.81 | #235 |
| 11 | Kimi K2.7 Code | 0.7 | #89 |
| 12 | MiMo-V2.5-Pro | 0.67 | #158 |
| 13 | Qwen 3.7 Max | 0.6 | #75 |
| 14 | Kimi K2.6 | 0.1 | #104 |
| 15 | Grok 4.3 | 0 | #188 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=age-of-llm · How It Works · Data refreshed daily, snapshot 2026-10-20.