DuelLab Overall — leaderboard
DuelLab evaluates model-generated game-playing programs by compiling submitted code and running head-to-head tournaments on hidden abstract strategy games.
Metric: Avg score (self-reported). Source: benchmarklist.com. Status: saturation imminent. 51 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 74.4 |
| 2 | GPT-5.4 | 74.2 |
| 3 | GPT-5.5 | 66.5 |
| 4 | GPT-5.2 | 66.1 |
| 5 | Claude Opus 4.6 | 65.2 |
| 6 | DeepSeek V4 Pro | 65 |
| 7 | GLM-5.1 | 63.1 |
| 8 | GPT-5.3 Codex | 62.4 |
| 9 | Gemini 3.1 Pro (Preview) | 62.3 |
| 10 | Claude Sonnet 4.6 | 60.9 |
| 11 | Qwen 3.6 Plus | 58.6 |
| 12 | GPT-5.4 Nano | 55.6 |
| 13 | GLM-5 | 55.5 |
| 14 | MiMo-V2.5-Pro | 54.2 |
| 15 | DeepSeek V4 Flash | 52.5 |
Interactive version: theaggregate.ai/benchmark?slug=duellab-overall · How the rankings work · Data refreshed daily, snapshot 2026-07-22.