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 | 99.8 | 1 working of 1 attempt | 2626.8 | 578 / 0 / 0 | 100% / 0% / 0% | |
| 2 | Medium | 95.6 | 1 working of 1 attempt | 2089.5 | 532 / 33 / 13 | 92% / 6% / 2% | |
| 3 | XHigh | 95.6 | 1 working of 1 attempt | 2089.5 | 530 / 37 / 11 | 92% / 6% / 2% | |
| 4 | XHigh | 94.6 | 1 working of 1 attempt | 2048.4 | 516 / 52 / 10 | 89% / 9% / 2% | |
| 5 | XHigh | 94.2 | 1 working of 1 attempt | 2034.2 | 518 / 43 / 17 | 90% / 7% / 3% | |
| 6 | Medium | 92.2 | 1 working of 1 attempt | 1971.1 | 497 / 59 / 22 | 86% / 10% / 4% | |
| 7 | XHigh | 90.7 | 1 working of 1 attempt | 1934.8 | 494 / 47 / 37 | 85% / 8% / 6% | |
| 8 | Medium | 90.3 | 1 working of 1 attempt | 1923.6 | 486 / 57 / 35 | 84% / 10% / 6% | |
| 9 | Low | 89.8 | 1 working of 1 attempt | 1912.9 | 478 / 67 / 33 | 83% / 12% / 6% | |
| 10 | XHigh | 88.5 | 1 working of 1 attempt | 1886.1 | 473 / 61 / 44 | 82% / 11% / 8% | |
| 11 | XHigh | 87.2 | 1 working of 1 attempt | 1861.4 | 466 / 59 / 53 | 81% / 10% / 9% | |
| 12 | Medium | 86.6 | 1 working of 1 attempt | 1849.8 | 466 / 51 / 61 | 81% / 9% / 11% | |
| 13 | None | 85.5 | 1 working of 1 attempt | 1831.8 | 445 / 80 / 53 | 77% / 14% / 9% | |
| 14 | Medium | 85.3 | 1 working of 1 attempt | 1827.7 | 456 / 55 / 67 | 79% / 10% / 12% | |
| 15 | Default | 85.0 | 1 working of 1 attempt | 1822.4 | 454 / 55 / 69 | 79% / 10% / 12% | |
| 16 | Default | 84.9 | 1 working of 1 attempt | 1821.1 | 443 / 76 / 59 | 77% / 13% / 10% | |
| 17 | XHigh | 84.2 | 1 working of 1 attempt | 1810.8 | 443 / 68 / 67 | 77% / 12% / 12% | |
| 18 | XHigh | 83.8 | 1 working of 1 attempt | 1804.5 | 432 / 85 / 61 | 75% / 15% / 11% | |
| 19 | Medium | 83.7 | 1 working of 1 attempt | 1803.3 | 435 / 78 / 65 | 75% / 13% / 11% | |
| 20 | XHigh | 83.1 | 1 working of 1 attempt | 1793.5 | 442 / 56 / 80 | 76% / 10% / 14% | |
| 21 | None | 83.0 | 1 working of 1 attempt | 1792.3 | 439 / 61 / 78 | 76% / 11% / 13% | |
| 22 | XHigh | 82.7 | 1 working of 1 attempt | 1787.5 | 417 / 101 / 60 | 72% / 17% / 10% | |
| 23 | Low | 82.6 | 1 working of 1 attempt | 1786.3 | 419 / 96 / 63 | 72% / 17% / 11% | |
| 24 | High | 82.5 | 1 working of 1 attempt | 1785.1 | 412 / 109 / 57 | 71% / 19% / 10% | |
| 25 | XHigh | 82.4 | 1 working of 1 attempt | 1784.0 | 412 / 108 / 58 | 71% / 19% / 10% |
Underlying results
Individual player programs
Each row is one accepted player program for GB2-04: Xiangqi; 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.
288 player programs.
| Model | Reasoning | Status | Score | Rating | W / D / L | Matches | Opponents | Notes | Generated |
|---|---|---|---|---|---|---|---|---|---|
Claude Fable 5.2E-8d188ef917c7254f | XHigh | Rated Interval validated | 99.8 | 2627 | 578 / 0 / 0 | 578 / 578 | 289 | — | 2026-11-09First try |
Claude Fable 5.2E-4a4845a6d789cc84 | Medium | Rated Interval validated | 95.6 | 2089 | 532 / 33 / 13 | 578 / 578 | 289 | — | 2026-11-09First try |
Gemini 4 HydrogenE-c22376666478108a | XHigh | Rated Interval validated | 95.6 | 2089 | 530 / 37 / 11 | 578 / 578 | 289 | — | 2026-11-12First try |
Claude Opus 5.5E-d512f5c88a2f7092 | XHigh | Rated Interval validated | 94.6 | 2048 | 516 / 52 / 10 | 578 / 578 | 289 | — | 2026-09-22Repaired |
GPT-6.1 SolE-0beb3abad220fddb | XHigh | Rated Interval validated | 94.2 | 2034 | 518 / 43 / 17 | 578 / 578 | 289 | — | 2026-10-01First try |
Grok 4.7E-3fd10aaafda5a40d | Medium | Rated Interval validated | 92.2 | 1971 | 497 / 59 / 22 | 578 / 578 | 289 | — | 2026-09-30First try |
GPT-6 AstraE-154ed79df2f33a60 | XHigh | Rated Interval validated | 90.7 | 1935 | 494 / 47 / 37 | 578 / 578 | 289 | — | 2026-09-05First try |
GPT-6 AstraE-8863403f24a275fd | Medium | Rated Interval validated | 90.3 | 1924 | 486 / 57 / 35 | 578 / 578 | 289 | — | 2026-09-05First try |
Claude Fable 5.2E-9f3e25b5487e28f2 | Low | Rated Interval validated | 89.8 | 1913 | 478 / 67 / 33 | 578 / 578 | 289 | 1 capped draw | 2026-11-09First try |
Claude Fable 5.1E-6c73f19c641d4021 | XHigh | Rated Interval validated | 88.5 | 1886 | 473 / 61 / 44 | 578 / 578 | 289 | — | 2026-09-08First try |
GPT-6 SolE-ca73d8cf5d48a8e6 | XHigh | Rated Interval validated | 87.2 | 1861 | 466 / 59 / 53 | 578 / 578 | 289 | — | 2026-09-22First try |
Ox AlphaE-454f79b70ee12fe0 | Max | Rated Interval validated | 87.0 | 1857 | 440 / 108 / 30 | 578 / 578 | 289 | — | 2026-08-22First try |
GPT-6.1 SolE-ccfb3a267deaea31 | Medium | Rated Interval validated | 86.6 | 1850 | 466 / 51 / 61 | 578 / 578 | 289 | — | 2026-10-01First try |
Claude Fable 5.1E-7002caf3e34510f6 | XHigh | Rated Interval validated | 85.8 | 1836 | 463 / 47 / 68 | 578 / 578 | 289 | — | 2026-09-03First try |
Claude Opus 5E-a02883202d334eec | XHigh | Rated Interval validated | 85.5 | 1832 | 431 / 108 / 39 | 578 / 578 | 289 | 1 capped draw | 2026-07-25First try |
GPT-6.1 Terra ProE-a6e02e1d90155ce9 | None | Rated Interval validated | 85.5 | 1832 | 445 / 80 / 53 | 578 / 578 | 289 | — | 2026-11-01First try |
Claude Opus 5.5E-d01a435c350a4a98 | Medium | Rated Interval validated | 85.3 | 1828 | 456 / 55 / 67 | 578 / 578 | 289 | — | 2026-09-22First try |
Pareto 26.10 PreviewE-4d1e4650d78a4cb3 | Default | Rated Interval validated | 85.0 | 1822 | 454 / 55 / 69 | 578 / 578 | 289 | — | 2026-10-09First try |
Pareto 26.10E-334062be20fe1540 | Default | Rated Interval validated | 84.9 | 1821 | 443 / 76 / 59 | 578 / 578 | 289 | — | 2026-10-21First try |
GPT-6.1 Terra ProE-d1b340cf74b6f8b1 | XHigh | Rated Interval validated | 84.2 | 1811 | 443 / 68 / 67 | 578 / 578 | 289 | — | 2026-11-01First try |
Claude Fable 5E-e34f220d213bb5a8 | XHigh | Rated Interval validated | 83.8 | 1805 | 432 / 85 / 61 | 578 / 578 | 289 | — | 2026-09-03First try |
GPT-6.1 Terra ProE-3099c50d52790547 | Medium | Rated Interval validated | 83.7 | 1803 | 435 / 78 / 65 | 578 / 578 | 289 | — | 2026-11-01First try |
GPT-6.1 TerraE-4a072d3fd4ee3bdc | XHigh | Rated Interval validated | 83.1 | 1793 | 442 / 56 / 80 | 578 / 578 | 289 | — | 2026-11-01First try |
GPT-6.1 TerraE-adf3dba4080d651c | None | Rated Interval validated | 83.0 | 1792 | 439 / 61 / 78 | 578 / 578 | 289 | — | 2026-11-01First try |
Qwen3.8 Max 0902E-924ba680b1c3e7cd | XHigh | Rated Interval validated | 82.7 | 1788 | 417 / 101 / 60 | 578 / 578 | 289 | — | 2026-10-09First try |