SUDOKU BENCHMARK · LOCAL, NO THINKING

Qwen3.5 9B

Alibaba · open weights · released 2026-02-23 · route qwen3.5:9b · Model profile · Chess results

Results

Score
095%: 0 to 20 · 7 of 42 grids so far
Grids solved
0 of 7
Blank cells filled correctly
9.3%
Wrong cells / empty cells / changed a given / no grid
0 / 0 / 4 / 3
Broke the rules (complete grid repeating a digit)
0
Near misses (three cells or fewer wrong)
0
Average stated confidence
0%
Average confidence when it had not solved it
0%
Sure but wrong (90% or more, unsolved)
0
Calibration (Brier, lower is better)
0.000
API-equivalent cost per grid
n/a
API-equivalent cost per solved grid
n/a
Seconds per grid
242
Tokens per grid (in / out)
1,732 / 2,402
Reasoning tokens per grid
n/a
Retries for an unreadable reply
7
Aborted on our side (not counted)
0

By tier

TierNeedsSolvedOutput tokensSeconds
CozySingles only0 of 13,000342
GentleScanning0 of 13,000357
SteadyLocked candidates0 of 13,000318
TrickyHidden pairs and triples0 of 13,000278
DeviousWings and fish0 of 11,606206
ExpertDeep wings and fish0 of 11,60566
BeyondNo human technique finishes it0 of 11,605123

Cells by the technique they needed

Each blank cell is tagged with the simplest technique that fills it, given everything a solver could fill before it.

TechniqueCellsFilled correctly
Last free cell425.0%
Scanning12314.6%
Locked candidates128.3%
Hidden pairs and triples10.0%
Wings and fish2222.7%
Deep wings and fish3429.4%
Search333.0%

Every answer