Skip to content
decisionbench
OverallSudokuTriageRobustness

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-04expert-01 →

Hard #05 sudoku race

27 clues · seed 3005 · 12 models · suite v1

12 AI decision models raced to solve this hard sudoku (27 clues), one digit at a time. Jev 1.14 by TypeSafe came first with 7 mistakes.

0.0s / 83.9s0/12 done

Results

#ModelMistakesCallsCostSolve timeStatus
1Jev 1.14TypeSafe723$0.0051956.4sSolved
2GPT-6 Luna DecisionsOpenAI1227$0.015911sSolved
3d1 UltraLiquid1426$0.0031348.7sSolved
4Clef FlashCloudflare1429$0.00541319sSolved
5Mercury DecideInception1528$0.00347714sSolved
6Decision-1Microsoft1629$0.00385417sSolved
7Clef OmniCloudflare1732$0.032613sSolved
8ClefCloudflare1731$0.011523sSolved
9Solar Decide FlashUpstage1732$0.045154sSolved
10d1Liquid2133$0.00206510sSolved
11Decider Flash v1.2Perplexity2134$0.00744155sSolved
12Decider v1.2Perplexity2639$0.00323184sSolved