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-05expert-02 →

Expert #01 sudoku race

23 clues · seed 4001 · 19 models · suite v1

19 AI decision models raced to solve this expert sudoku (23 clues), one digit at a time. d1 Ultra by Liquid came first with 10 mistakes.

0.0s / 84.9s0/19 done

Results

#ModelMistakesCallsCostSolve timeStatus
1d1 UltraLiquid1026$0.00316621sSolved
2Span-02Respan1531$0.00560771sSolved
3Jev 1.13TypeSafeolder1630$0.0071726.5sSolved
4Drex v2.0Nace.AI1832$0.00721424sSolved
5Clef FlashCloudflare1834$0.00740027sSolved
6Clef OmniCloudflare2034$0.035212sSolved
7d1Liquid2134$0.00196311sSolved
8Decision-1Microsoft2133$0.00438919sSolved
9Jev 1.14TypeSafe2238$0.00958711sSolved
10Kev 4BJared Palmer2337$0.00499626sSolved
11Drex v1.5Nace.AIolder2439$0.00506838sSolved
12Decider v1.2Perplexity2440$0.00331685sSolved
13Mercury DecideInception2536$0.00460121sSolved
14Decider Flash v1.2Perplexity2640$0.00878164sSolved
15GPT-6 Luna DecisionsOpenAI2743$0.029020sSolved
16Solar Decide FlashUpstage2943$0.067477sSolved
17Tev1 4B ExperimentalTogether3043$0.045150sSolved
18Classify-1ScaleDown3045$0.055252sSolved
19ClefCloudflare3143$0.015133sSolved