OPENROUTER_OPENAI_GPT-5-MINI · 20260725-codex-smoke2

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

openrouter_openai_gpt-5-mini · 1st of 1 compare all models →

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

score
—
intensity
—
nausea
—
best round
—
rated rounds
0/1
tokens / cost
2.0M in / 13k out · $0.10
1 round(s)

Round 1 build failed trace (102) →

build failed (7/7 pieces placed): finish_and_test: set_status(testing): Track is not a complete circuit [track starts at {"bank":0,"dir":1,"slope":0,"x":50,"y":50,"z":112} and currently ends at {"bank":0,"dir":0,"slope":0,"x":52,"y":51,"z":112}; a closed circuit requires these to match]

track program (7 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "end_station",
    "right_turn_5",
    "right_turn_5",
    "right_turn_5"
  ],
  "ride_type": 51,
  "start": {
    "dir": 1,
    "x": 50,
    "y": 50
  }
}
park screenshot, openrouter_openai_gpt-5-mini round 1

← 20260725-codex-smoke2 · all runs