MICRO MOUSE BENCHMARK · REASONING MEDIUM

Claude Haiku 5.5

Anthropic · released 2026-10-07 · route claude-haiku-5-5 · Model profile · Chess results · Sudoku results · Mines results · Nonograms results · FreeCell results · Starjar results · Nations Royale results · Wildhaven results · Hustle Sim results · Elemental Armory results · Ravelin results · Minecraft results

Results

Score
5395%: 39 to 67
Campaign score against the bot
6349 to 78
Daily mazes faster than the median person
0 of 6
Reached the dock
14 of 14 mazes
Faster than the bot
0 of 14
Reached / out of time / out of replies
14 / 0 / 0
Share of the route covered, on average
100%
Wall hits per run
8.4
Drift boosts / pad boosts per run
0.0 / 0.2
Replies per run
49.5
Control segments per reply
2.09
Values moved into range
126
Segments cut at the one-second limit
16
Lines that were not segments
0
Replies with no readable segment
0.0%
Retries for an unreadable reply
0
Model minutes per run
22.3
Tokens per reply (in / out)
2,603 / 5,569
Reasoning tokens per reply
5,479
API-equivalent cost per run
$0.152
Seconds per reply
27.0
Replies cut off at the length limit
0.0%
Aborted on our side (not counted)
55

By tier

TierAgainstRunsScoreReachedPaceWall hits / runTest pilot
Daily mazesmedian person6536 of 6539.848
Workbenchbot2642 of 2649.052
Cold Storagebot2662 of 2663.549
Fast Lanebot2652 of 2657.550
Night Shiftbot2602 of 2609.050

Every run