OPENROUTER_MOONSHOTAI_KIMI-K3 · 20260725-codex-kimi

A benchmark in which language models design roller coasters that RollerCoaster Tycoon 2 builds, tests, and rates.

openrouter_moonshotai_kimi-k3 · 1st of 1 compare all models →

design + open note · steel twister · codex · design mode — models build from scratch

score
0.57
intensity
0.20
nausea
0.10
best round
1
rated rounds
1/1
similarity
0.43
tokens / cost
$1.23
1 round(s)

Round 1 trace (77) →

excitement 0.57 intensity 0.20 nausea 0.10 length 31 drops 2 airtime 0 tested ok similarity 0.43 (nearest: Penguin Toboggan)
track program (21 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "middle_station",
    "end_station",
    "flat_to_up_25",
    "up_25",
    "up_25",
    "up_25",
    "up_25_to_flat",
    "left_turn_5",
    "left_turn_5",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    "flat",
    "s_bend_left",
    "left_turn_5",
    "left_turn_3"
  ],
  "ride_type": 51,
  "start": {
    "dir": 2,
    "x": 40,
    "y": 40
  }
}
park screenshot, openrouter_moonshotai_kimi-k3 round 1

← 20260725-codex-kimi · all runs