SUDOKU BENCHMARK · REASONING MEDIUM

Gemini 4 Neon

Google · released 2026-11-18 · route gemini-4-neon · Model profile · Chess results

Results

Score
10095%: 84 to 100 · 14 of 42 grids so far
Grids solved
14 of 14
Blank cells filled correctly
100.0%
Wrong cells / empty cells / changed a given / no grid
0 / 0 / 0 / 0
Broke the rules (complete grid repeating a digit)
0
Near misses (three cells or fewer wrong)
0
Average stated confidence
100%
Average confidence when it had not solved it
n/a
Sure but wrong (90% or more, unsolved)
0
Calibration (Brier, lower is better)
0.000
API-equivalent cost per grid
$0.299
API-equivalent cost per solved grid
$0.299
Seconds per grid
136
Tokens per grid (in / out)
2,289 / 19,454
Reasoning tokens per grid
19,346
Retries for an unreadable reply
0
Aborted on our side (not counted)
0

By tier

TierNeedsSolvedOutput tokensSeconds
CozySingles only2 of 29,83669
GentleScanning2 of 211,92279
SteadyLocked candidates2 of 214,503102
TrickyHidden pairs and triples2 of 225,730188
DeviousWings and fish2 of 222,134153
ExpertDeep wings and fish2 of 221,882157
BeyondNo human technique finishes it2 of 230,168200

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 cell121100.0%
Scanning250100.0%
Locked candidates88100.0%
Hidden pairs and triples104100.0%
Wings and fish63100.0%
Deep wings and fish54100.0%
Search81100.0%

Every answer