Benchmarks / Reasoning
BoardBench
Models play 30 unseen puzzles across ten board games (Xiangqi, chess, Go, backgammon, shogi, Scrabble, Kalah, Gomoku, Othello, Connect Four). An engine scores each legal move from 0 to 1 and any illegal move 0; the value is the mean score × 100.
BoardBench
Higher is better · best effort per model
full table below
| # | Model | Effort | Score |
|---|---|---|---|
| 1 | Gemini 4 | high | 66.3% |
| 2 | GPT-6.1 Sol | high | 56.4% |
| 3 | GLM 5.3 Flash | high | 43.9% |
| 4 | GPT-6 Luna | high | 42.5% |
| 5 | DeepSeek V4.1 Flash | high | 39.9% |
| 6 | Haiku 5.5 | high | 39.0% |
| 7 | Opus 5.5 | high | 37.6% |