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.
Results
| # | Model | Mistakes | Calls | Cost | Solve time | Status |
|---|
| 1 | d1 UltraLiquid | 5 | 22 | $0.002449 | 7.3s | Solved |
| 2 | Drex v2.0Nace.AI | 5 | 21 | $0.004033 | 17s | Solved |
| 3 | ClefCloudflare | 9 | 23 | $0.0108 | 19s | Solved |
| 4 | d1Liquid | 10 | 26 | $0.001350 | 7.3s | Solved |
| 5 | Jev 1.14TypeSafe | 11 | 25 | $0.005610 | 7.5s | Solved |
| 6 | Clef OmniCloudflare | 11 | 25 | $0.0245 | 11s | Solved |
| 7 | Decider Flash v1.2Perplexity | 11 | 26 | $0.004848 | 40s | Solved |
| 8 | Jev 1.13TypeSafeolder | 12 | 26 | $0.006100 | 9.4s | Solved |
| 9 | Mercury DecideInception | 13 | 29 | $0.003618 | 16s | Solved |
| 10 | Clef FlashCloudflare | 13 | 28 | $0.005346 | 21s | Solved |
| 11 | Classify-1ScaleDown | 13 | 27 | $0.0270 | 26s | Solved |
| 12 | Solar Decide FlashUpstage | 13 | 28 | $0.0374 | 42s | Solved |
| 13 | GPT-6 Luna DecisionsOpenAI | 17 | 30 | $0.0187 | 14s | Solved |
| 14 | Span-02Respan | 17 | 32 | $0.005928 | 71s | Solved |
| 15 | Kev 4BJared Palmer | 17 | 31 | $0.003904 | 85s | Solved |
| 16 | Decision-1Microsoft | 19 | 32 | $0.004028 | 18s | Solved |
| 17 | Decider v1.2Perplexity | 19 | 35 | $0.002769 | 37s | Solved |
| 18 | Tev1 4B ExperimentalTogether | 19 | 33 | $0.0278 | 45s | Solved |
| 19 | Drex v1.5Nace.AIolder | 21 | 36 | $0.004264 | 32s | Solved |