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
← hard-02hard-04 →

Hard #03 sudoku race

27 clues · seed 3003 · 19 models · suite v1

19 AI decision models raced to solve this hard sudoku (27 clues), one digit at a time. d1 Ultra by Liquid came first with 5 mistakes.

0.0s / 84.6s0/19 done

Results

#ModelMistakesCallsCostSolve timeStatus
1d1 UltraLiquid522$0.0024497.3sSolved
2Drex v2.0Nace.AI521$0.00403317sSolved
3ClefCloudflare923$0.010819sSolved
4d1Liquid1026$0.0013507.3sSolved
5Jev 1.14TypeSafe1125$0.0056107.5sSolved
6Clef OmniCloudflare1125$0.024511sSolved
7Decider Flash v1.2Perplexity1126$0.00484840sSolved
8Jev 1.13TypeSafeolder1226$0.0061009.4sSolved
9Mercury DecideInception1329$0.00361816sSolved
10Clef FlashCloudflare1328$0.00534621sSolved
11Classify-1ScaleDown1327$0.027026sSolved
12Solar Decide FlashUpstage1328$0.037442sSolved
13GPT-6 Luna DecisionsOpenAI1730$0.018714sSolved
14Span-02Respan1732$0.00592871sSolved
15Kev 4BJared Palmer1731$0.00390485sSolved
16Decision-1Microsoft1932$0.00402818sSolved
17Decider v1.2Perplexity1935$0.00276937sSolved
18Tev1 4B ExperimentalTogether1933$0.027845sSolved
19Drex v1.5Nace.AIolder2136$0.00426432sSolved