StratMem-Bench: leaderboard
Metric: Strategic memory coverage SMC (%): share of virtual-character responses that use every must-use memory and appropriate nice-to-use memories, over all instances weighted by scenario counts; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 51.45 |
| 2 | Gemini 3 Pro (Preview) | 50.68 |
| 3 | Llama 4 Maverick | 50.23 |
| 4 | GPT-5.2 | 48.55 |
| 5 | Qwen 3 235B A22B 2507 Instruct | 47.18 |
| 6 | Qwen 3 Max | 46.42 |
| 7 | GPT-5 Chat | 46.27 |
| 8 | DeepSeek V3.2 | 45.66 |
Interactive version: theaggregate.ai/benchmark?slug=stratmem-bench · How It Works · Data refreshed daily, snapshot 2026-10-07.