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 | 99.6 | 1 working of 1 attempt | 2546.1 | 559 / 3 / 0 | 99% / 1% / 0% | |
| 2 | XHigh | 96.6 | 1 working of 1 attempt | 2157.7 | 521 / 37 / 4 | 93% / 7% / 1% | |
| 3 | XHigh | 96.0 | 1 working of 1 attempt | 2123.3 | 518 / 35 / 9 | 92% / 6% / 2% | |
| 4 | High | 96.0 | 1 working of 1 attempt | 2123.3 | 525 / 21 / 16 | 93% / 4% / 3% | |
| 5 | XHigh | 95.6 | 1 working of 1 attempt | 2104.0 | 517 / 32 / 13 | 92% / 6% / 2% | |
| 6 | XHigh | 95.1 | 1 working of 1 attempt | 2082.5 | 516 / 28 / 18 | 92% / 5% / 3% | |
| 7 | XHigh | 94.8 | 1 working of 1 attempt | 2065.8 | 510 / 35 / 17 | 91% / 6% / 3% | |
| 8 | Medium | 94.3 | 1 working of 1 attempt | 2047.1 | 511 / 27 / 24 | 91% / 5% / 4% | |
| 9 | XHigh | 94.1 | 1 working of 1 attempt | 2038.1 | 499 / 48 / 15 | 89% / 9% / 3% | |
| 10 | High | 93.9 | 1 working of 1 attempt | 2032.4 | 491 / 62 / 9 | 87% / 11% / 2% | |
| 11 | Medium | 93.8 | 1 working of 1 attempt | 2029.5 | 506 / 31 / 25 | 90% / 6% / 4% | |
| 12 | Medium | 91.1 | 1 working of 1 attempt | 1946.7 | 468 / 73 / 21 | 83% / 13% / 4% | |
| 13 | Default | 90.6 | 1 working of 1 attempt | 1934.3 | 477 / 49 / 36 | 85% / 9% / 6% | |
| 14 | Medium | 90.3 | 1 working of 1 attempt | 1926.3 | 477 / 45 / 40 | 85% / 8% / 7% | |
| 15 | XHigh | 88.7 | 1 working of 1 attempt | 1891.1 | 461 / 58 / 43 | 82% / 10% / 8% | |
| 16 | XHigh | 87.7 | 1 working of 1 attempt | 1870.8 | 453 / 62 / 47 | 81% / 11% / 8% | |
| 17 | Low | 86.7 | 1 working of 1 attempt | 1850.2 | 440 / 75 / 47 | 78% / 13% / 8% | |
| 18 | XHigh | 82.6 | 2 working of 2 attempts | 1789.6 program range 1727–1852 | 794 / 226 / 104 | 71% / 20% / 9% | |
| 19 | XHigh | 82.0 | 1 working of 1 attempt | 1774.1 | 367 / 165 / 30 | 65% / 29% / 5% | |
| 20 | Max | 82.0 | 1 working of 1 attempt | 1774.1 | 407 / 85 / 70 | 72% / 15% / 12% | |
| 21 | Low | 81.8 | 1 working of 1 attempt | 1771.6 | 383 / 131 / 48 | 68% / 23% / 9% | |
| 22 | XHigh | 81.7 | 1 working of 1 attempt | 1770.4 | 402 / 92 / 68 | 72% / 16% / 12% | |
| 23 | Low | 81.0 | 1 working of 1 attempt | 1760.9 | 371 / 146 / 45 | 66% / 26% / 8% | |
| 24 | Medium | 80.8 | 1 working of 1 attempt | 1757.4 | 351 / 183 / 28 | 62% / 33% / 5% | |
| 25 | Max | 79.7 | 1 working of 1 attempt | 1742.7 | 377 / 118 / 67 | 67% / 21% / 12% |
Underlying results
Individual player programs
Each row is one accepted player program for GB2-02: Chess; 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.
280 player programs.
| Model | Reasoning | Status | Score | Rating | W / D / L | Matches | Opponents | Notes | Generated |
|---|---|---|---|---|---|---|---|---|---|
Claude Fable 5.2E-56812c272561d614 | XHigh | Rated Interval validated | 99.6 | 2546 | 559 / 3 / 0 | 562 / 562 | 281 | — | 2026-11-09First try |
Gemini 4 HydrogenE-e82ed6b9c68ae82b | XHigh | Rated Interval validated | 96.6 | 2158 | 521 / 37 / 4 | 562 / 562 | 281 | — | 2026-11-12First try |
Claude Sonnet 5.5E-25c4c1579034e616 | XHigh | Rated Interval validated | 96.0 | 2123 | 518 / 35 / 9 | 562 / 562 | 281 | — | 2026-09-30First try |
Gemini 4 CarbonE-3a5ab7a8b8473dea | High | Rated Interval validated | 96.0 | 2123 | 525 / 21 / 16 | 562 / 562 | 281 | — | 2026-11-12Repaired |
Claude Opus 5.5E-8da2a605b97075fa | XHigh | Rated Interval validated | 95.6 | 2104 | 517 / 32 / 13 | 562 / 562 | 281 | — | 2026-09-22First try |
Claude Fable 5.1E-f86f2f46e2c71b4a | XHigh | Rated Interval validated | 95.1 | 2082 | 516 / 28 / 18 | 562 / 562 | 281 | — | 2026-09-08First try |
GPT-6 AstraE-c91ee6859ab8877e | XHigh | Rated Interval validated | 94.8 | 2066 | 510 / 35 / 17 | 562 / 562 | 281 | — | 2026-09-05First try |
Claude Opus 5.5E-4692bc19babebbb1 | Medium | Rated Interval validated | 94.3 | 2047 | 511 / 27 / 24 | 562 / 562 | 281 | — | 2026-09-22Repaired |
GPT-6.1 SolE-755187947384baf1 | XHigh | Rated Interval validated | 94.1 | 2038 | 499 / 48 / 15 | 562 / 562 | 281 | — | 2026-10-01First try |
Gemini 4 ArgonE-35eb3b86b76e7c9f | High | Rated Interval validated | 93.9 | 2032 | 491 / 62 / 9 | 562 / 562 | 281 | — | 2026-10-21Repaired |
Claude Fable 5.2E-692854324abe471c | Medium | Rated Interval validated | 93.8 | 2030 | 506 / 31 / 25 | 562 / 562 | 281 | — | 2026-11-09First try |
Gemini 4 CarbonE-207a6d6cbd1d627a | Medium | Rated Interval validated | 91.1 | 1947 | 468 / 73 / 21 | 562 / 562 | 281 | — | 2026-11-12First try |
Pareto 26.10 PreviewE-e13381315ea87cb7 | Default | Rated Interval validated | 90.6 | 1934 | 477 / 49 / 36 | 562 / 562 | 281 | — | 2026-10-09First try |
Claude Fable 5.1E-f4d4ba39a6f2270e | Medium | Rated Interval validated | 90.3 | 1926 | 477 / 45 / 40 | 562 / 562 | 281 | — | 2026-09-03First try |
GPT-6.1 Terra ProE-26c8994965e09404 | XHigh | Rated Interval validated | 88.7 | 1891 | 461 / 58 / 43 | 562 / 563 | 281 | 1 fault | 2026-11-01First try |
Grok 4.7E-ba03a6f2135ea911 | XHigh | Rated Interval validated | 87.7 | 1871 | 453 / 62 / 47 | 562 / 562 | 281 | — | 2026-09-30First try |
Claude Opus 5E-fcdd98d74e80134a | XHigh | Rated Interval validated | 86.7 | 1852 | 455 / 46 / 61 | 562 / 562 | 281 | — | 2026-07-25First try |
Gemini 4 CarbonE-6a27e144b3514939 | Low | Rated Interval validated | 86.7 | 1850 | 440 / 75 / 47 | 562 / 562 | 281 | — | 2026-11-12First try |
GLM-5.3 FlashX LegalE-b4431d08cc503c1e | Max | Rated Interval validated | 82.0 | 1774 | 407 / 85 / 70 | 562 / 562 | 281 | — | 2026-11-03First try |
GPT-6 SolE-6b0e5cea9d2e9834 | XHigh | Rated Interval validated | 82.0 | 1774 | 367 / 165 / 30 | 562 / 562 | 281 | — | 2026-09-22First try |
Kimi K3.1 ProE-313ef2e01b0b5aa0 | Low | Rated Interval validated | 81.8 | 1772 | 383 / 131 / 48 | 562 / 562 | 281 | — | 2026-10-24First try |
Grok 4.6E-892d7d5eef11e96c | XHigh | Rated Interval validated | 81.7 | 1770 | 402 / 92 / 68 | 562 / 562 | 281 | — | 2026-09-03First try |
Claude Fable 5.1E-85e618bbeae3eac9 | Low | Rated Interval validated | 81.0 | 1761 | 371 / 146 / 45 | 562 / 562 | 281 | — | 2026-09-03First try |
GPT-6 AstraE-db1b6aef6e22ec35 | Medium | Rated Interval validated | 80.8 | 1757 | 351 / 183 / 28 | 562 / 562 | 281 | — | 2026-09-05First try |
DeepSeek V4.1 Flash 1017E-5e795839a2bce09c | Max | Rated Interval validated | 79.7 | 1743 | 377 / 118 / 67 | 562 / 562 | 281 | — | 2026-10-20First try |