Benchmark
LLM Chess Wrong Moves per 1000 Moves
Lower scores are better.
- Models measured
- 149
- Scores recorded
- 206
- Sources
- 1
- IQ weight
- 1.97
Score distribution
- 43.7
- Mean
- 9.1
- Median
- 76.9
- Std dev
- 0.483
- 10th percentile
- 141
- 90th percentile
- 0
- Min
This benchmark has a weight of 1.97 for chat models, measured over 149 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 201 | Gemma 3 27B ITnone@iq4_xs· LLM Chess Leaderboard data docs | 285.5 |
| 202 | Gemma 3 4B ITnone@q8_0· LLM Chess Leaderboard data docs | 287.8 |
| 203 | Qwen3 32Bdefault· LLM Chess Leaderboard data docs | 325.5 |
| 204 | GLM 4.7 Flashdefault@q4· LLM Chess Leaderboard data docs | 339.3 |
| 205 | Mercury Coder Smallnone· LLM Chess Leaderboard data docs | 343.1 |
| 206 | Qwen3.5 0.8Bnone@bf16· LLM Chess Leaderboard data docs | 458.3 |
Showing 201–206 of 206 results.
Data credits: LLM Chess by Maxim Saplin, Apache-2.0 (Copyright [2024] [Maxim Saplin]) (licence)