RUN 20260813-qwen38-design (design)

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

Standings

design mode — models build from scratch

modelscoreintensitynausea similaritybesttokens / cost
1st openrouter_qwen_qwen3.8-2.4t-a95b 7.30 8.88 4.68 0.31 round 2/6 383k in / 1.3M out · $28.40

score = excitement, scaled down when similarity to a stock library design exceeds 0.5 (an exact or mirrored copy scores 0)

openrouter_qwen_qwen3.8-2.4t-a95bfull detail →
excitementintensity×build failed
0515intensity ceiling 10excitementintensityR1R2×round 3: build failed (0/0 pieces placed): finish_and_test: no active ride; call new_ride firstR3R4R5×round 6: build failed (0/0 pieces placed): tools/call: http://127.0.0.1:8791/mcp?lease=openrouter_qwen_qwen3.8-2.4t-a95b-r6-1786631663: Connection Failed: Connect error: Connection refused (os error 61)R6

Round 1 verified circuit (56 pieces) trace (240, 5 rejected) →

excitement6.19 intensity6.44 nausea3.24
length 440 m drops 5 airtime 17 tested ok similarity 0.32 (nearest: Doubledrop)
boosterboosterleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downbrakes
56 pieces · 8 on the lift · 6 inversions
track program (56 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "end_station",
    {
      "speed": 26,
      "t": "booster"
    },
    {
      "chain": true,
      "t": "flat_to_up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25_to_flat"
    },
    "flat",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "flat",
    "flat",
    {
      "speed": 28,
      "t": "booster"
    },
    "left_corkscrew_up",
    "right_corkscrew_down",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "right_turn_5",
    "s_bend_right",
    "s_bend_right",
    "s_bend_right",
    "s_bend_right",
    "flat_to_up_25",
    "up_25",
    "up_25_to_flat",
    "flat_to_right_bank",
    "right_helix_down_small",
    "right_helix_down_small",
    "right_bank_to_flat",
    {
      "speed": 12,
      "t": "brakes"
    },
    "flat",
    "right_turn_5"
  ],
  "ride_type": 51,
  "start": {
    "dir": 2,
    "x": 50,
    "y": 60
  }
}

download scored .park

replay

Round 2 verified circuit (124 pieces) trace (244, 1 rejected) →

excitement7.30 intensity8.88 nausea4.68
length 848 m drops 15 airtime 84 tested ok similarity 0.31 (nearest: To The Death)
boosterboosterleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downboosterleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downleft_corkscrew_upright_corkscrew_downboosterbrakesbrakesbrakes
124 pieces · 9 on the lift · 12 inversions · 74 added, 6 removed vs the round before (50 of 56 pieces kept)
track program (124 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "end_station",
    {
      "speed": 28,
      "t": "booster"
    },
    {
      "chain": true,
      "t": "flat_to_up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25_to_flat"
    },
    "flat",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    {
      "speed": 30,
      "t": "booster"
    },
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat",
    "flat",
    "flat",
    "flat",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "flat",
    "flat",
    "flat",
    "flat",
    "s_bend_right",
    "s_bend_left",
    "flat",
    "flat",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "flat",
    "flat",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat",
    "flat",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    {
      "speed": 27,
      "t": "booster"
    },
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat",
    "flat",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "left_corkscrew_up",
    "right_corkscrew_down",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "s_bend_right",
    "s_bend_right",
    "s_bend_right",
    "s_bend_right",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "s_bend_right",
    "s_bend_left",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "s_bend_right",
    "s_bend_left",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    {
      "speed": 27,
      "t": "booster"
    },
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "s_bend_right",
    "s_bend_left",
    "flat",
    "flat",
    "flat",
    "flat",
    "flat",
    "flat_to_right_bank",
    "banked_right_turn_5",
    "right_bank_to_flat",
    "flat",
    "flat",
    {
      "speed": 12,
      "t": "brakes"
    },
    {
      "speed": 12,
      "t": "brakes"
    },
    {
      "speed": 8,
      "t": "brakes"
    },
    "flat",
    "flat",
    "flat"
  ],
  "ride_type": 51,
  "start": {
    "dir": 2,
    "x": 40,
    "y": 92
  }
}

download scored .park

replay

Round 3 build failed trace (4, 1 rejected) →

build failed (0/0 pieces placed): finish_and_test: no active ride; call new_ride first

track program (0 pieces)
{
  "pieces": [],
  "ride_type": 51,
  "start": {
    "dir": 0,
    "x": 0,
    "y": 0
  }
}

download scored .park

park screenshot, openrouter_qwen_qwen3.8-2.4t-a95b round 3

Round 4 verified circuit (30 pieces) trace (44, 2 rejected) →

excitement1.22 intensity0.62 nausea0.31
length 187 m drops 1 airtime 0 tested ok similarity 0.57 (nearest: Doubledrop) penalized score 1.06
brakesbrakes
30 pieces · 6 on the lift
track program (30 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "end_station",
    {
      "chain": true,
      "t": "flat_to_up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    "up_25_to_flat",
    "flat",
    "left_turn_5",
    "left_turn_5",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    "flat",
    "flat",
    {
      "speed": 8,
      "t": "brakes"
    },
    "flat",
    "flat",
    "flat",
    "left_turn_5",
    "left_turn_5",
    {
      "speed": 4,
      "t": "brakes"
    }
  ],
  "ride_type": 51,
  "start": {
    "dir": 2,
    "x": 30,
    "y": 38
  }
}

download scored .park

replay

Round 5 verified circuit (66 pieces) trace (411, 18 rejected) →

excitement5.86 intensity4.75 nausea2.29
length 486 m drops 4 airtime 24 tested ok similarity 0.38 (nearest: Radical Air)
boosterleft_vertical_loop
66 pieces · 7 on the lift · 1 inversion · 40 added, 4 removed vs the round before (26 of 30 pieces kept)
track program (66 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "end_station",
    {
      "chain": true,
      "t": "flat_to_up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25_to_flat"
    },
    "flat",
    "left_turn_5",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    {
      "speed": 30,
      "t": "booster"
    },
    "flat",
    "flat_to_up_25",
    "left_vertical_loop",
    "down_25_to_flat",
    "flat",
    "flat",
    "s_bend_left",
    "s_bend_right",
    "flat",
    "flat_to_up_25",
    "up_25_to_flat",
    "flat_to_down_25",
    "down_25_to_flat",
    "flat",
    "left_turn_5",
    "flat",
    "flat",
    "flat",
    "flat_to_up_25",
    "up_25",
    "up_25_to_flat",
    "flat",
    "flat_to_down_25",
    "down_25",
    "down_25_to_flat",
    "flat",
    "flat",
    "flat",
    "flat",
    "left_turn_5",
    "flat",
    "flat",
    "s_bend_left",
    "s_bend_right",
    "s_bend_left",
    "s_bend_right",
    "s_bend_left",
    "s_bend_right",
    "s_bend_left",
    "s_bend_right",
    "flat",
    "left_turn_5",
    "flat",
    "flat",
    "flat"
  ],
  "ride_type": 51,
  "start": {
    "dir": 2,
    "x": 30,
    "y": 44
  }
}

download scored .park

replay

Round 6 build failed trace (177, 4 rejected) →

build failed (0/0 pieces placed): tools/call: http://127.0.0.1:8791/mcp?lease=openrouter_qwen_qwen3.8-2.4t-a95b-r6-1786631663: Connection Failed: Connect error: Connection refused (os error 61)

track program (0 pieces)
{
  "pieces": [],
  "ride_type": 51,
  "start": {
    "dir": 0,
    "x": 0,
    "y": 0
  }
}
no capture

← all runs