NONOGRAMS BENCHMARK · LOCAL, NO THINKING

Gemma 3 4B

Google · open weights · released 2025-03-12 · route gemma3:4b · Model profile · Chess results · Sudoku results · Mines results · FreeCell results · Starjar results · Nations Royale results · Wildhaven results · Hustle Sim results · Elemental Armory results · Micro Mouse results · Ravelin results · Minecraft results

Results

Score
095%: 0 to 22 · 6 of 36 base puzzles so far
Puzzles solved
0 of 6
Cells that match the solution
0.0%
Rows and columns whose clue it satisfies
0.0%
Wrong cells / left cells unknown / no grid
0 / 0 / 6
Near misses (three cells or fewer wrong)
0
Average stated confidence
95%
Average confidence when it had not solved it
95%
Sure but wrong (90% or more, unsolved)
5
Calibration (Brier, lower is better)
0.902
API-equivalent cost per solved puzzle
n/a
Tokens per puzzle (in / out)
1,380 / 1,198
Reasoning tokens per puzzle
n/a
Retries for an unreadable reply
6
API-equivalent cost per puzzle
n/a
Seconds per puzzle
80.9
Replies cut off at the length limit
0.0%
Aborted on our side (not counted)
0

By tier

TierNeedsSolvedOutput tokensSeconds
5 by 5Line logic, a round or two0 of 110617.6
10 by 10Line logic, about 3 rounds0 of 120823.9
15 by 15Line logic, about 5 rounds0 of 12,68582.6
20 by 20Line logic, about 8 rounds0 of 1016.5
25 by 25Line logic, about 11 rounds0 of 13,354312.8
BeyondLine logic stalls; needs contradiction or search0 of 183231.7

Cells by the depth of logic they need

Each cell is tagged with the round of line logic that settles it. Beyond line logic means only contradiction or search does.

DepthCellsRight
First pass4030.0%
Rounds 2-36130.0%
Rounds 4-73320.0%
Round 8 on1590.0%
Beyond line logic930.0%

Every puzzle