Benchmark menu

Per-game leaderboard

GB2-05: Hex

This page shows the per-game leaderboard for GB2-05: Hex in the xhigh variants. Entrants are ranked by their relative per-game score within this game.

GB2-05: Hex

Hex

Hex is a connection game on a hexagonal grid where players race to join opposite board edges, with a swap option after the first move.

Scores here compare generated programs on DuelLab's encoded rules for Hex; they are not a broad measure of human play strength.

Entrants

Every admitted unique code identity for GB2-05: Hex. Capped draws are rated; unilateral model faults are listed separately.

Showing 28 entrants.
EntrantModelStatusScoreRatingW / L / DRated / attemptedOpponentsCapped drawsModel faultsAdmitted pluginsRepairGenerated
E-5aa4402ccce7b855Claude Fable 5Rated92.92106187 / 5 / 0192 / 19296001Repaired2026-07-02
E-7e9bc3ebc12ef500Claude Opus 4.5Rated57.61650112 / 80 / 0192 / 19296001First pass2026-07-02
E-6774d700e60de31aClaude Opus 4.8Rated51.91577114 / 79 / 0193 / 19397001Repaired2026-07-02
E-c902807aab849f31Claude Sonnet 5Rated63.61728129 / 64 / 0193 / 19397001First pass2026-07-03
E-b306c8f5ffbbec38DeepSeek V4 FlashRated12.9107430 / 162 / 0192 / 19296001Repaired2026-07-02
E-b3d889b0631db85fDeepSeek V4 FlashRated59.51675123 / 69 / 0192 / 19296001Repaired2026-07-02
E-0ca4798825ceaccbDeepSeek V4 ProRated85.72013176 / 16 / 0192 / 19296001Repaired2026-07-02
E-a72dd193e1dc4331Gemini 3.1 Flash LiteRated6.498928 / 164 / 0192 / 19296001Repaired2026-07-02
E-b50da126631c1ea1Gemini 3.5 FlashRated54.31608115 / 77 / 0192 / 19296001Repaired2026-07-02
E-c5ad048b6fb4e93dGPT 5Rated82.31969169 / 23 / 0192 / 19296001First pass2026-07-02
E-f6403f7a30e19e82GPT-5.4Rated100.02198188 / 4 / 0192 / 19296001First pass2026-07-02
E-481582e8f8e3217eGPT-5.4 MiniRated70.61818146 / 46 / 0192 / 19296001First pass2026-07-02
E-95d1a872f51b78e7GPT-5.4 MiniRated43.51468101 / 91 / 0192 / 19296001First pass2026-07-02
E-54b9ca4996a5589dGPT-5.4 NanoRated74.91874144 / 48 / 0192 / 19296001First pass2026-07-02
E-b05a7e0987de9967GPT-5.4 NanoRated87.22032175 / 17 / 0192 / 19296001First pass2026-07-09
E-762e9d80a60b41edGPT-5.5Rated76.61896163 / 29 / 0192 / 19296001First pass2026-07-02
E-25a83ddaeb1268a4GPT-5.6 LunaRated57.01643128 / 64 / 0192 / 19296001First pass2026-07-10
E-af9c2c2998294ad0GPT-5.6 SolRated79.91939168 / 24 / 0192 / 19296001First pass2026-07-10
E-012ecb9ce0709c05GPT-5.6 TerraRated79.71936168 / 24 / 0192 / 19296001Repaired2026-07-10
E-1954b83d439b455cMistral Medium 3.5Rated23.9121645 / 147 / 0192 / 19296001Repaired2026-07-02
E-48ed5c5e00ef38fbNemotron 3 Ultra 550B A55BRated31.7131652 / 140 / 0192 / 19296001First pass2026-07-02
E-da3b98638ff53ea9Nemotron 3 Ultra 550B A55BRated47.3151799 / 97 / 0196 / 200100001Repaired2026-07-11
E-4fc62dc65156d576O3Rated60.21684128 / 64 / 0192 / 19296001First pass2026-07-03
E-358338a9fbdbcb6aStep 3.7 FlashRated50.0155390 / 102 / 0192 / 19296001Repaired2026-07-02
E-92dc0ac61d798488Hy3 PreviewRated63.91731135 / 57 / 0192 / 19296001Repaired2026-07-03
E-235e4beedf22bc99GLM-5.2Rated57.71651140 / 56 / 0196 / 200100001Repaired2026-07-11
E-503ecd021099d8f9GLM-5.2Rated41.3144083 / 110 / 0193 / 19397001Repaired2026-07-02
E-51c8f79c1671819eGLM-5.2Rated57.01643125 / 67 / 0192 / 19296001Repaired2026-07-02

GB2-05: Hex leaderboard

Entrants are ranked by relative per-game score (0–100). Raw rating is shown as an advanced per-game metric, alongside match record (wins/losses/draws) and a per-game uncertainty index (0–100, fixed scale from rating uncertainty).

Reasoning level: XHigh Game: GB2-05: Hex

Scroll sideways to see every column.

GB2-05: Hex - XHigh variants
Rank Model Score Contributing programs Raw Elo W / L / D Uncertainty
1GPT-5.4100.012197.6188/4/00.0
2Claude Fable 592.912105.6187/5/00.0
3DeepSeek V4 Pro85.712012.7176/16/00.0
4GPT 582.311969.2169/23/00.0
5GPT-5.4 Nano81.021952.8319/65/00.0
6GPT-5.6 Sol79.911938.7168/24/00.0
7GPT-5.6 Terra79.711935.9168/24/00.0
8GPT-5.576.611895.8163/29/00.0
9Hy3 Preview63.911731.4135/57/00.0
10Claude Sonnet 563.611727.8129/64/00.0
11O360.211684.2128/64/00.0
12Claude Opus 4.557.611650.3112/80/00.0
13GPT-5.4 Mini57.021643.2247/137/00.0
14GPT-5.6 Luna57.011643.0128/64/00.0
15Gemini 3.5 Flash54.311607.8115/77/00.0
16GLM-5.252.031578.1348/233/00.0
17Claude Opus 4.851.911576.7114/79/00.0
18Step 3.7 Flash50.011552.690/102/00.0
19Nemotron 3 Ultra 550B A55B39.521416.4151/237/00.0
20DeepSeek V4 Flash36.221374.2153/231/00.0
21Mistral Medium 3.523.911215.945/147/00.0
22Gemini 3.1 Flash Lite6.41988.828/164/00.0