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.

GB2-06: International Draughts - Reasoning Variants
Rank Model Reasoning Score Programs Raw rating Combined record (W / D / L)
1Claude Fable 5.2XHigh99.71 working of 1 attempt3017.2584 / 0 / 0
2Claude Opus 5.5XHigh99.41 working of 1 attempt2827.0579 / 2 / 3
3Gemini 4 CarbonHigh99.11 working of 1 attempt2747.4575 / 5 / 4
4Gemini 4 HydrogenXHigh99.01 working of 1 attempt2733.6577 / 0 / 7
5Claude Fable 5.1XHigh98.01 working of 1 attempt2578.9567 / 5 / 12
6Gemini 4 ArgonHigh96.61 working of 1 attempt2463.2554 / 13 / 17
7Claude Fable 5.2Low93.61 working of 1 attempt2309.0529 / 24 / 31
8Pareto 26.10Default91.71 working of 1 attempt2244.3517 / 25 / 42
9Claude Haiku 5.5XHigh91.31 working of 1 attempt2231.6513 / 28 / 43
10GPT-6.1 Terra ProMedium91.31 working of 1 attempt2231.6515 / 24 / 45
11Kimi K3.1 ProLow90.91 working of 1 attempt2219.3516 / 17 / 51
12Claude Opus 5.5Medium90.51 working of 1 attempt2205.1512 / 19 / 53
13GLM-5.3High90.51 working of 1 attempt2205.1507 / 29 / 48
14GPT-6.1 TerraMedium90.41 working of 1 attempt2202.8499 / 44 / 41
15GPT-6.1 TerraXHigh90.41 working of 1 attempt2202.8511 / 20 / 53
16GPT-6.1 Terra ProXHigh90.41 working of 1 attempt2202.8503 / 36 / 45
17Kimi K3.1 ProHigh90.31 working of 1 attempt2200.4505 / 31 / 48
18Claude Sonnet 5.5XHigh90.21 working of 1 attempt2195.9499 / 41 / 44
19Gemini 4 ArgonMedium90.01 working of 1 attempt2191.3503 / 31 / 50
20GPT-6.1 SolMedium89.71 working of 1 attempt2182.4504 / 25 / 55
21Nex N2.6Low89.41 working of 1 attempt2173.6501 / 27 / 56
22Gemini 4 CarbonMedium89.01 working of 1 attempt2165.0504 / 17 / 63
23GLM-5.3Max89.01 working of 1 attempt2165.0493 / 39 / 52
24Grok 4.7XHigh88.91 working of 1 attempt2160.7505 / 13 / 66
25Claude Fable 5.2Medium88.41 working of 1 attempt2148.2496 / 25 / 63

Individual player programs

Each row is one accepted player program for GB2-06: International Draughts; 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.

292 player programs.

ModelReasoningStatusScoreRatingW / D / LMatchesOpponentsNotesGenerated
Claude Fable 5.2E-24b55f22d86d7a18XHighRated Interval validated99.73017584 / 0 / 0584 / 584292—2026-11-09First try
Claude Opus 5.5E-53134a8de9454448XHighRated Interval validated99.42827579 / 2 / 3584 / 584292—2026-09-22First try
Gemini 4 CarbonE-e9aac1a0e7f7caa7HighRated Interval validated99.12747575 / 5 / 4584 / 584292—2026-11-12First try
Gemini 4 HydrogenE-48bcbced49818817XHighRated Interval validated99.02734577 / 0 / 7584 / 584292—2026-11-12First try
Claude Fable 5.1E-bba525f4f6adfed4XHighRated Interval validated98.02579567 / 5 / 12584 / 5842921 capped draw2026-09-08First try
Gemini 4 ArgonE-ec915524f0712927HighRated Interval validated96.62463554 / 13 / 17584 / 584292—2026-10-21Repaired
Claude Fable 5.2E-75b7eb3d343354e6LowRated Interval validated93.62309529 / 24 / 31584 / 5842929 capped draws2026-11-09First try
Pareto 26.10E-5d01b5fd5968de69DefaultRated Interval validated91.72244517 / 25 / 42584 / 5842923 capped draws2026-10-21First try
Claude Haiku 5.5E-7508ffbcb30ce9d7XHighRated Interval validated91.32232513 / 28 / 43584 / 5842926 capped draws2026-10-09First try
GPT-6.1 Terra ProE-e1b1e3f73d88104cMediumRated Interval validated91.32232515 / 24 / 45584 / 5842924 capped draws2026-11-01First try
Kimi K3.1 ProE-65169852dc5bea55LowRated Interval validated90.92219516 / 17 / 51584 / 5842926 capped draws2026-10-24First try
Claude Opus 5.5E-cb36bcb6ba10266bMediumRated Interval validated90.52205512 / 19 / 53584 / 5842924 capped draws2026-09-22First try
GLM-5.3E-256f22a4f85c8ce5HighRated Interval validated90.52205507 / 29 / 48584 / 5842926 capped draws2026-09-03First try
GPT-6.1 TerraE-312f19e7cdaaeff5XHighRated Interval validated90.42203511 / 20 / 53584 / 5842926 capped draws2026-11-01First try
GPT-6.1 TerraE-c4542a0a341b2146MediumRated Interval validated90.42203499 / 44 / 41584 / 58429213 capped draws2026-11-01First try
GPT-6.1 Terra ProE-4b426f92084ee8caXHighRated Interval validated90.42203503 / 36 / 45584 / 5842929 capped draws2026-11-01First try
Kimi K3.1 ProE-02f59a1e4530ae4dHighRated Interval validated90.32200505 / 31 / 48584 / 5842927 capped draws2026-10-24First try
Claude Sonnet 5.5E-33fb3ee55eb8f02fXHighRated Interval validated90.22196499 / 41 / 44584 / 584292—2026-09-30First try
Gemini 4 ArgonE-670f358c134cd627MediumRated Interval validated90.02191503 / 31 / 50584 / 5842927 capped draws2026-10-21First try
GPT-6.1 SolE-a4495c1439ca4ab3MediumRated Interval validated89.72182504 / 25 / 55584 / 5842925 capped draws2026-10-01First try
Nex N2.6E-1f2335b1989b8170LowRated Interval validated89.42174501 / 27 / 56584 / 5842924 capped draws2026-11-12First try
Gemini 4 CarbonE-fc544518b11cd5deMediumRated Interval validated89.02165504 / 17 / 63584 / 5842921 capped draw2026-11-12First try
GLM-5.3E-206af7e7704349b5MaxRated Interval validated89.02165493 / 39 / 52584 / 584292—2026-10-09First try
Grok 4.7E-c0e64450a9e18936XHighRated Interval validated88.92161505 / 13 / 66584 / 5842921 capped draw2026-09-30Repaired
Gemini 3.6 FlashE-905531e52a8460f5HighRated Interval validated88.52150506 / 6 / 72584 / 5842922 capped draws2026-07-24Repaired