Leaderboard
By model and setting
Each model and reasoning setting appears once. If a model produced more than one player program for this game, the row stands for all of them: Programs counts the working attempts out of the counted attempts, Score and Raw rating average the programs (weighted by attempts; the small range is the lowest and highest program rating), and the combined record adds up their matches. Score is relative within this game (0–100). Scores here are working-program strength: failed generation attempts are left out and count as 0 only in a model's headline score.
| Rank | Model | Reasoning | Score | Programs | Raw rating | Combined record (W / D / L) | W / D / L % |
|---|---|---|---|---|---|---|---|
| 1 | XHigh | 98.4 | 1 working of 1 attempt | 2502.3 | 559 / 4 / 7 | 98% / 1% / 1% | |
| 2 | XHigh | 98.3 | 1 working of 1 attempt | 2493.5 | 557 / 11 / 4 | 97% / 2% / 1% | |
| 3 | High | 98.0 | 1 working of 1 attempt | 2454.0 | 401 / 3 / 8 | 97% / 1% / 2% | |
| 4 | XHigh | 97.7 | 1 working of 1 attempt | 2422.4 | 558 / 1 / 13 | 98% / <1% / 2% | |
| 5 | XHigh | 97.6 | 1 working of 1 attempt | 2415.6 | 556 / 2 / 14 | 97% / <1% / 2% | |
| 6 | XHigh | 97.2 | 1 working of 1 attempt | 2376.6 | 406 / 5 / 9 | 97% / 1% / 2% | |
| 7 | Default | 96.4 | 1 working of 1 attempt | 2311.6 | 546 / 3 / 21 | 96% / 1% / 4% | |
| 8 | Medium | 96.2 | 1 working of 1 attempt | 2296.9 | 547 / 7 / 22 | 95% / 1% / 4% | |
| 9 | Medium | 96.0 | 1 working of 1 attempt | 2281.5 | 543 / 4 / 25 | 95% / 1% / 4% | |
| 10 | XHigh | 94.8 | 1 working of 1 attempt | 2209.3 | 538 / 0 / 34 | 94% / 0% / 6% | |
| 11 | XHigh | 94.6 | 1 working of 1 attempt | 2199.9 | 529 / 10 / 31 | 93% / 2% / 5% | |
| 12 | XHigh | 94.6 | 1 working of 1 attempt | 2198.5 | 536 / 2 / 34 | 94% / <1% / 6% | |
| 13 | Low | 94.3 | 1 working of 1 attempt | 2181.5 | 532 / 2 / 36 | 93% / <1% / 6% | |
| 14 | Medium | 93.9 | 1 working of 1 attempt | 2163.7 | 528 / 8 / 36 | 92% / 1% / 6% | |
| 15 | XHigh | 93.3 | 2 working of 2 attempts | 2135.3 program range 2105–2166 | 1048 / 13 / 85 | 91% / 1% / 7% | |
| 16 | XHigh | 93.2 | 1 working of 1 attempt | 2130.2 | 523 / 6 / 43 | 91% / 1% / 8% | |
| 17 | Medium | 93.0 | 1 working of 1 attempt | 2118.9 | 504 / 43 / 23 | 88% / 8% / 4% | |
| 18 | High | 92.1 | 1 working of 1 attempt | 2081.7 | 493 / 50 / 27 | 86% / 9% / 5% | |
| 19 | Default | 92.1 | 1 working of 1 attempt | 2081.5 | 515 / 7 / 50 | 90% / 1% / 9% | |
| 20 | Medium | 92.1 | 1 working of 1 attempt | 2081.0 | 506 / 28 / 38 | 88% / 5% / 7% | |
| 21 | Low | 91.6 | 1 working of 1 attempt | 2061.0 | 515 / 3 / 56 | 90% / 1% / 10% | |
| 22 | Medium | 91.0 | 1 working of 1 attempt | 2040.7 | 505 / 15 / 52 | 88% / 3% / 9% | |
| 23 | XHigh | 89.0 | 1 working of 1 attempt | 1972.1 | 462 / 65 / 43 | 81% / 11% / 8% | |
| 24 | Low | 87.6 | 1 working of 1 attempt | 1933.9 | 351 / 7 / 54 | 85% / 2% / 13% | |
| 25 | None | 87.6 | 1 working of 1 attempt | 1932.6 | 483 / 8 / 79 | 85% / 1% / 14% |
Underlying results
Individual player programs
Each row is one accepted player program for GB2-03: Shogi; a model can have more than one. Score is the 0–100 result in this game, and its rating and W / D / L record show how it played. Scores are working-program strength: failed generation attempts have no row and count only in a model's headline score.
290 player programs.
| Model | Reasoning | Status | Score | Rating | W / D / L | Matches | Opponents | Notes | Generated |
|---|---|---|---|---|---|---|---|---|---|
Claude Fable 5.2E-53600b7725dfeb1b | XHigh | Rated Interval validated | 98.4 | 2502 | 559 / 4 / 7 | 570 / 572 | 285 | — | 2026-11-09First try |
GPT-6 AstraE-18e6aa627f4b1222 | XHigh | Rated Interval validated | 98.3 | 2494 | 557 / 11 / 4 | 572 / 574 | 286 | — | 2026-09-05First try |
Gemini 4 CarbonE-06596dc7382fab4f | High | Rated Interval validated | 98.0 | 2454 | 401 / 3 / 8 | 412 / 414 | 206 | — | 2026-11-12First try |
GPT-6.1 SolE-27bd4aa164eef135 | XHigh | Rated Interval validated | 97.7 | 2422 | 558 / 1 / 13 | 572 / 573 | 286 | — | 2026-10-01First try |
Claude Opus 5.5E-0ac738bb34b1e7c8 | XHigh | Rated Interval validated | 97.6 | 2416 | 556 / 2 / 14 | 572 / 574 | 286 | — | 2026-09-22First try |
Gemini 4 HydrogenE-a064debfcb33dc65 | XHigh | Rated Interval validated | 97.2 | 2377 | 406 / 5 / 9 | 420 / 422 | 210 | — | 2026-11-12Repaired |
Pareto 26.10E-3f0521308d4debcb | Default | Rated Interval validated | 96.4 | 2312 | 546 / 3 / 21 | 570 / 572 | 285 | — | 2026-10-21First try |
Gemini 4 CarbonE-1e3ddf2284bebdfd | Medium | Rated Interval validated | 96.2 | 2297 | 547 / 7 / 22 | 576 / 578 | 288 | — | 2026-11-12First try |
GPT-6.1 SolE-fb177ebd31849b47 | Medium | Rated Interval validated | 96.0 | 2281 | 543 / 4 / 25 | 572 / 574 | 286 | — | 2026-10-01First try |
Claude Sonnet 5.5E-4a7793fa8a4833be | XHigh | Rated Interval validated | 94.8 | 2209 | 538 / 0 / 34 | 572 / 574 | 286 | — | 2026-09-30First try |
GPT-6.1 Terra ProE-f5fa770829216201 | XHigh | Rated Interval validated | 94.6 | 2200 | 529 / 10 / 31 | 570 / 572 | 285 | — | 2026-11-01First try |
Claude Fable 5.1E-b53393b05159fc95 | XHigh | Rated Interval validated | 94.6 | 2198 | 536 / 2 / 34 | 572 / 574 | 286 | — | 2026-09-08First try |
Claude Fable 5.2E-c1d0091908115770 | Low | Rated Interval validated | 94.3 | 2182 | 532 / 2 / 36 | 570 / 572 | 285 | — | 2026-11-09First try |
Claude Fable 5.1E-9f7c077c66189b69 | XHigh | Rated Interval validated | 94.0 | 2169 | 531 / 4 / 37 | 572 / 574 | 286 | — | 2026-09-03First try |
Claude Opus 5E-7caa6c5081d57676 | XHigh | Rated Interval validated | 94.0 | 2166 | 529 / 8 / 37 | 574 / 576 | 287 | — | 2026-07-25First try |
GPT-6 AstraE-4633465f951f8cae | Medium | Rated Interval validated | 93.9 | 2164 | 528 / 8 / 36 | 572 / 573 | 286 | — | 2026-09-05First try |
GPT-6 SolE-049711eb0d28bfa8 | XHigh | Rated Interval validated | 93.2 | 2130 | 523 / 6 / 43 | 572 / 574 | 286 | — | 2026-09-22First try |
Claude Fable 5.2E-be0fb5f0eb971c6e | Medium | Rated Interval validated | 93.0 | 2119 | 504 / 43 / 23 | 570 / 571 | 285 | — | 2026-11-09First try |
Claude Opus 5E-38759212c025e8eb | XHigh | Rated Interval validated | 92.7 | 2105 | 519 / 5 / 48 | 572 / 574 | 286 | — | 2026-07-25First try |
Gemini 4 ArgonE-674c6806d0094ec2 | High | Rated Interval validated | 92.1 | 2082 | 493 / 50 / 27 | 570 / 572 | 285 | — | 2026-10-21Repaired |
Pareto 26.10 PreviewE-74f4f6139b879936 | Default | Rated Interval validated | 92.1 | 2082 | 515 / 7 / 50 | 572 / 574 | 286 | — | 2026-10-09First try |
Claude Opus 5.5E-8b93ed37e8717e57 | Medium | Rated Interval validated | 92.1 | 2081 | 506 / 28 / 38 | 572 / 574 | 286 | — | 2026-09-22First try |
Claude Fable 5.1E-529bc866427d95d5 | Low | Rated Interval validated | 91.6 | 2061 | 515 / 3 / 56 | 574 / 576 | 287 | — | 2026-09-03First try |
Claude Fable 5.1E-5c6dabbae6a319cf | Medium | Rated Interval validated | 91.0 | 2041 | 505 / 15 / 52 | 572 / 574 | 286 | — | 2026-09-03First try |
GPT-6.1 TerraE-f53bdd0e9e794a81 | XHigh | Rated Interval validated | 89.0 | 1972 | 462 / 65 / 43 | 570 / 571 | 285 | — | 2026-11-01First try |