Seed 2
The seed fixes every random outcome in the game: mazes, cactus, sunflowers, pumpkins and each drone's random(). Every model here played the same farm, so the differences are the agents'.
race
Score over game time, every model
GPT-6.1 Sol75.9Claude Opus 5.575.5Claude Fable 5.172.5GPT-6 Astra63.2Claude Sonnet 5.560.8Grok 4.757.4Claude Haiku 5.537.2GPT-6 Luna27.5
Data table
| Game time | GPT-6.1 Sol | Claude Opus 5.5 | Claude Fable 5.1 | GPT-6 Astra | Claude Sonnet 5.5 | Grok 4.7 | Claude Haiku 5.5 | GPT-6 Luna |
|---|---|---|---|---|---|---|---|---|
| 0s | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
| 1h 00m | 24.3 | 31.9 | 26.6 | 21.3 | 20.7 | 31.7 | 17.4 | 3.84 |
| 2h 00m | 42.3 | 49.0 | 40.9 | 34.4 | 38.9 | 46.8 | 28.4 | 11.3 |
| 3h 00m | 54.5 | 57.6 | 54.7 | 47.6 | 45.6 | 51.7 | 32.5 | 13.1 |
| 4h 00m | 64.8 | 67.2 | 63.5 | 56.3 | 53.3 | 54.7 | 34.9 | 17.5 |
| 5h 00m | 66.7 | 72.4 | 69.7 | 60.0 | 57.8 | 56.4 | 36.3 | 18.7 |
| 6h 00m | 75.9 | 75.5 | 72.5 | 63.2 | 60.8 | 57.4 | 37.2 | 27.5 |
One line per model, coloured by vendor; a vendor's other models are dashed. Hover to compare at any moment.
standings
Standings on this seed
- 1GPT-6.1 SolCodex · medium75.87
Hay 8.91 points
Wood 9.33 points
Carrot 8.57 points
Pumpkin 7.16 points
Cactus 6.88 points
Power 5.15 points
Bone 6.24 points
Weird substance 8.42 points
Gold 5.81 points
Water 5.31 points
Fertilizer 4.09 points
- 2Claude Opus 5.5Claude Code · medium75.53
- Hay 8.59 points
- Wood 9.34 points
- Carrot 8.45 points
- Pumpkin 8.15 points
- Cactus 8.45 points
- Power 4.22 points
- Bone 6.53 points
- Weird substance 5.80 points
- Gold 6.42 points
- Water 5.47 points
- Fertilizer 4.11 points
- 3Claude Fable 5.1Claude Code · medium72.54
- Hay 8.56 points
- Wood 9.40 points
- Carrot 9.25 points
- Pumpkin 7.12 points
- Cactus 5.61 points
- Power 3.95 points
- Bone 6.25 points
- Weird substance 8.26 points
- Gold 4.65 points
- Water 5.42 points
- Fertilizer 4.08 points
- 4GPT-6 AstraCodex · medium63.16
- Hay 6.53 points
- Wood 6.94 points
- Carrot 6.51 points
- Pumpkin 6.01 points
- Cactus 6.16 points
- Power 4.44 points
- Bone 6.27 points
- Weird substance 6.45 points
- Gold 4.95 points
- Water 4.88 points
- Fertilizer 4.01 points
- 5Claude Sonnet 5.5Claude Code · medium60.75
- Hay 7.13 points
- Wood 7.04 points
- Carrot 6.93 points
- Pumpkin 6.88 points
- Cactus 6.40 points
- Power 5.13 points
- Bone 1.04 points
- Weird substance 5.70 points
- Gold 5.57 points
- Water 4.96 points
- Fertilizer 3.97 points
- 6Grok 4.7Cursor · medium57.38
- Hay 6.48 points
- Wood 7.35 points
- Carrot 6.51 points
- Pumpkin 4.05 points
- Cactus 4.21 points
- Power 4.22 points
- Bone 4.79 points
- Weird substance 6.62 points
- Gold 4.04 points
- Water 4.98 points
- Fertilizer 4.14 points
- 7Claude Haiku 5.5Claude Code · medium37.24
- Hay 5.64 points
- Wood 5.59 points
- Carrot 5.59 points
- Pumpkin 5.61 points
- Cactus 2.51 points
- Power 4.26 points
- Bone not produced
- Weird substance not produced
- Gold not produced
- Water 4.33 points
- Fertilizer 3.71 points
- 8GPT-6 LunaCodex · medium27.46
- Hay 4.19 points
- Wood 4.50 points
- Carrot 3.58 points
- Pumpkin 3.81 points
- Cactus 3.61 points
- Power not produced
- Bone 4.55 points
- Weird substance not produced
- Gold not produced
- Water 3.20 points
- Fertilizer not produced
unlocks
When each model first produced each resource
| Model | | | | | | | | | | | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-6.1 Sol | 1m 00s | 9m 00s | 14m 50s | 43m 00s | 1h 41m | 1h 19m | 20m 10s | 30m 40s | 53m 10s | 1h 25m | 2h 55m |
| Claude Opus 5.5 | 1m 10s | 6m 20s | 12m 10s | 27m 00s | 1h 23m | 37m 30s | 19m 10s | 26m 00s | 29m 30s | 40m 10s | 1h 28m |
| Claude Fable 5.1 | 1m 40s | 6m 00s | 12m 00s | 28m 40s | 1h 12m | 35m 30s | 14m 10s | 28m 10s | 31m 00s | 1h 03m | 2h 01m |
| GPT-6 Astra | 1m 00s | 10m 00s | 16m 20s | 30m 10s | 2h 00m | 1h 27m | 26m 30s | 50m 30s | 51m 00s | 1h 47m | 3h 45m |
| Claude Sonnet 5.5 | 1m 10s | 9m 40s | 16m 30s | 31m 50s | 1h 33m | 53m 40s | 20m 10s | 45m 50s | 1h 44m | 1h 49m | 2h 27m |
| Grok 4.7 | 3m 10s | 12m 30s | 19m 30s | 55m 20s | 1h 07m | 39m 40s | 20m 30s | 27m 40s | 28m 40s | 44m 40s | 1h 11m |
| Claude Haiku 5.5 | 1m 30s | 10m 30s | 27m 50s | 49m 30s | 1h 41m | 46m 10s | 31m 30s | 1h 32m | · | · | · |
| GPT-6 Luna | 2m 20s | 1h 06m | 1h 31m | 3h 01m | 5h 29m | · | 1h 33m | · | · | · | 5h 34m |
Game time of the first 10-second sample with any of the resource. A dot: never produced.
how-they-played
How each model played this seed
| Model | Resources opened | Study | Action overhead | Median stint | Cost per point |
|---|---|---|---|---|---|
| GPT-6.1 Sol | 11 of 11 | 25 | 7% | 10m 30s | $0.0085 |
| Claude Opus 5.5 | 11 of 11 | 34 | 12% | 3m 04s | $0.068 |
| Claude Fable 5.1 | 11 of 11 | 50 | 12% | 4m 17s | $0.298 |
| GPT-6 Astra | 11 of 11 | 20 | 5% | 15m 25s | $0.027 |
| Claude Sonnet 5.5 | 11 of 11 | 22 | 12% | 3m 40s | $0.072 |
| Grok 4.7 | 11 of 11 | 83 | 11% | 2m 20s | $0.162 |
| Claude Haiku 5.5 | 8 of 11 | 45 | 17% | 2m 00s | $0.030 |
| GPT-6 Luna | 7 of 11 | 66 | 10% | 2m 31s | $0.0067 |
Resources opened: scored resources produced at least once. Study: docs and research reads. Action overhead: game time spent looking, reading and typing (10 s per call). Median stint: how long a program usually ran before the next change. Cost per point: API-equivalent estimate ÷ score.