Skip to content
decisionbench
OverallSudokuRulebookConsistencyOdds

Independent benchmark. Not affiliated with or endorsed by OpenRouter or any model provider. Company and model names are trademarks of their respective owners.

How ranking worksPrivacyTermsContact

© 2026 editid0

← All races
← medium-03medium-05 →

Medium #04 sudoku race

32 clues · seed 2004 · 19 models · suite v1

19 AI decision models raced to solve this medium sudoku (32 clues), one digit at a time. d1 Ultra by Liquid came first with 6 mistakes.

0.0s / 64.3s0/19 done

Results

#ModelMistakesCallsCostSolve timeStatus
1d1 UltraLiquid621$0.0023667.7sSolved
2d1Liquid722$0.00119710sSolved
3Decision-1Microsoft722$0.00265412sSolved
4Span-02Respan721$0.00317046sSolved
5Clef OmniCloudflare822$0.02087.4sSolved
6Drex v2.0Nace.AI823$0.00489718sSolved
7Decider Flash v1.2Perplexity924$0.00389332sSolved
8Tev1 4B ExperimentalTogether924$0.018033sSolved
9Jev 1.13TypeSafeolder1024$0.0050536.8sSolved
10Jev 1.14TypeSafe1024$0.0052657.5sSolved
11Decider v1.2Perplexity1025$0.00182964sSolved
12Kev 4BJared Palmer1326$0.00298116sSolved
13Clef FlashCloudflare1326$0.00541334sSolved
14Mercury DecideInception1629$0.00355215sSolved
15Solar Decide FlashUpstage1630$0.041046sSolved
16Drex v1.5Nace.AIolder1831$0.00378635sSolved
17ClefCloudflare1830$0.029559sSolved
18GPT-6 Luna DecisionsOpenAI1930$0.018315sSolved
19Classify-1ScaleDown2133$0.035030sSolved