Chess APO Bench: leaderboard
Metric: Puzzle accuracy (%) on the 559-puzzle Lichess test split with the unoptimized base prompt (every position must match the single reference move), mean of three runs. Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.5 Flash (Medium) | 55.4 |
| 2 | GPT-5.6 Luna (Low) | 31.72 |
| 3 | DeepSeek V4 Pro (0813) | 14.49 |
| 4 | Claude Haiku 4.5 | 11.63 |
| 5 | Qwen 3.8 27B | 8.47 |
| 6 | GPT-4o Mini | 5.84 |
Interactive version: theaggregate.ai/benchmark?slug=chess-apo-bench · How It Works · Data refreshed daily, snapshot 2026-10-07.