MINES BENCHMARK · LOCAL, THINKING LOW

Qwen3.8 Flash-Next

Alibaba · open weights · released 2026-08-26 · route qwen3.8-flash-next · Model profile · Chess results · Sudoku results · FreeCell results · Hustle Sim results · Elemental Armory results · Micro Mouse results

Results

Score
10095%: 47 to 100 · 1 of 24 base boards so far
Boards won
1 of 1
Safe cells cleared beyond the opening
100.0%
Won / hit a mine / unfinished / unreadable
1 / 0 / 0 / 0
Ended for changing nothing five turns running
0
Reached the turn cap
0
Flags left on safe cells
0 in 0 boards
Illegal actions
0 of 49 (0.0%)
Wasted actions (already open or flagged)
9 of 49 (18.4%)
Unreadable replies
0
Turns per board
6.0
Requests per board
6.0
API-equivalent cost per win
n/a
Tokens per board (in / out)
3,767 / 34,216
Reasoning tokens per board
n/a
API-equivalent cost per board
n/a
Seconds per board
3,094.2
Replies cut off at the length limit
0.0%
Aborted on our side (not counted)
1

By tier

TierBoard, minesWonClearedTurnsOutput tokensSeconds
Beginner9×9, 101 of 1100%6.034,2163,094.2
Intermediate16×16, 400 of 0n/an/an/an/a
Expert30×16, 990 of 0n/an/an/an/a
Master30×24, 2000 of 0n/an/an/an/a

Cells by the technique that proves them

Each safe cell covered at the start is tagged with the simplest deduction that first proves it safe, given everything a perfect player could open before it.

TechniqueCellsOpened
Cascade from an empty cell33100.0%
One number26100.0%
Two numbers together3100.0%

Every board