Seed 3
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
Claude Opus 5.575.4GPT-6.1 Sol74.6Claude Fable 5.171.3GPT-6 Astra69.9Grok 4.758.0Claude Sonnet 5.549.2Claude Haiku 5.537.1GPT-6 Luna15.1
Data table
| Game time | Claude Opus 5.5 | GPT-6.1 Sol | Claude Fable 5.1 | GPT-6 Astra | Grok 4.7 | Claude Sonnet 5.5 | 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 | 37.3 | 22.1 | 26.1 | 22.8 | 15.0 | 23.3 | 16.7 | 3.83 |
| 2h 00m | 45.6 | 44.4 | 40.7 | 42.3 | 33.9 | 40.5 | 28.0 | 3.86 |
| 3h 00m | 62.0 | 50.4 | 49.7 | 48.1 | 47.2 | 45.5 | 32.4 | 4.64 |
| 4h 00m | 67.9 | 56.1 | 61.9 | 60.9 | 51.6 | 47.4 | 34.8 | 6.89 |
| 5h 00m | 73.3 | 67.8 | 67.2 | 65.4 | 55.2 | 48.5 | 36.2 | 9.64 |
| 6h 00m | 75.4 | 74.6 | 71.3 | 69.9 | 58.0 | 49.2 | 37.1 | 15.1 |
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
- 1Claude Opus 5.5Claude Code · medium75.45
Hay 8.36 points
Wood 9.22 points
Carrot 8.40 points
Pumpkin 7.50 points
Cactus 7.87 points
Power 4.69 points
Bone 6.86 points
Weird substance 6.16 points
Gold 6.78 points
Water 5.49 points
Fertilizer 4.11 points
- 2GPT-6.1 SolCodex · medium74.56
- Hay 8.25 points
- Wood 8.93 points
- Carrot 8.13 points
- Pumpkin 7.42 points
- Cactus 7.71 points
- Power 5.18 points
- Bone 6.71 points
- Weird substance 7.61 points
- Gold 5.37 points
- Water 5.14 points
- Fertilizer 4.11 points
- 3Claude Fable 5.1Claude Code · medium71.30
- Hay 7.34 points
- Wood 8.47 points
- Carrot 7.91 points
- Pumpkin 6.47 points
- Cactus 6.97 points
- Power 4.54 points
- Bone 6.64 points
- Weird substance 7.53 points
- Gold 6.09 points
- Water 5.34 points
- Fertilizer 3.99 points
- 4GPT-6 AstraCodex · medium69.91
- Hay 8.00 points
- Wood 8.40 points
- Carrot 7.71 points
- Pumpkin 6.57 points
- Cactus 6.90 points
- Power 4.03 points
- Bone 5.71 points
- Weird substance 8.06 points
- Gold 5.28 points
- Water 5.15 points
- Fertilizer 4.08 points
- 5Grok 4.7Cursor · medium57.98
- Hay 6.48 points
- Wood 7.19 points
- Carrot 5.90 points
- Pumpkin 4.71 points
- Cactus 5.76 points
- Power 4.38 points
- Bone 5.28 points
- Weird substance 4.72 points
- Gold 4.96 points
- Water 4.66 points
- Fertilizer 3.96 points
- 6Claude Sonnet 5.5Claude Code · medium49.17
- Hay 5.15 points
- Wood 5.08 points
- Carrot 4.99 points
- Pumpkin 4.96 points
- Cactus 4.57 points
- Power 2.78 points
- Bone 6.19 points
- Weird substance 3.72 points
- Gold 4.16 points
- Water 3.84 points
- Fertilizer 3.73 points
- 7Claude Haiku 5.5Claude Code · medium37.08
- Hay 5.27 points
- Wood 5.34 points
- Carrot 5.38 points
- Pumpkin 5.50 points
- Cactus 4.98 points
- Power 2.85 points
- Bone not produced
- Weird substance not produced
- Gold not produced
- Water 4.39 points
- Fertilizer 3.37 points
- 8GPT-6 LunaCodex · medium15.09
- Hay 3.89 points
- Wood 3.40 points
- Carrot 2.82 points
- Pumpkin 2.54 points
- Cactus not produced
- Power not produced
- Bone not produced
- Weird substance not produced
- Gold not produced
- Water 2.44 points
- Fertilizer not produced
unlocks
When each model first produced each resource
| Model | | | | | | | | | | | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude Opus 5.5 | 1m 10s | 7m 00s | 12m 50s | 34m 00s | 50m 10s | 2h 02m | 15m 30s | 25m 00s | 26m 20s | 43m 00s | 52m 40s |
| GPT-6.1 Sol | 1m 00s | 8m 50s | 20m 40s | 54m 10s | 1h 06m | 1h 38m | 20m 30s | 42m 30s | 1h 16m | 1h 26m | 4h 11m |
| Claude Fable 5.1 | 1m 40s | 15m 10s | 21m 30s | 32m 40s | 1h 55m | 52m 40s | 23m 10s | 36m 30s | 36m 50s | 1h 01m | 2h 04m |
| GPT-6 Astra | 1m 00s | 9m 50s | 14m 40s | 1h 09m | 1h 21m | 49m 00s | 14m 30s | 48m 10s | 48m 50s | 1h 31m | 3h 15m |
| Grok 4.7 | 2m 40s | 6m 50s | 24m 20s | 1h 04m | 2h 02m | 1h 26m | 35m 10s | 37m 50s | 1h 21m | 1h 29m | 2h 06m |
| Claude Sonnet 5.5 | 1m 10s | 4m 30s | 10m 40s | 21m 30s | 1h 10m | 35m 20s | 21m 00s | 31m 50s | 1h 21m | 1h 29m | 1h 58m |
| Claude Haiku 5.5 | 1m 40s | 11m 30s | 27m 40s | 59m 10s | 1h 56m | 51m 20s | 25m 30s | 1h 46m | · | · | · |
| GPT-6 Luna | 2m 10s | 2h 08m | 4h 43m | 5h 16m | · | · | 5h 14m | · | · | · | · |
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 |
|---|---|---|---|---|---|
| Claude Opus 5.5 | 11 of 11 | 35 | 11% | 2m 50s | $0.063 |
| GPT-6.1 Sol | 11 of 11 | 25 | 6% | 10m 40s | $0.0043 |
| Claude Fable 5.1 | 11 of 11 | 65 | 12% | 5m 00s | $0.198 |
| GPT-6 Astra | 11 of 11 | 19 | 5% | 10m 20s | $0.027 |
| Grok 4.7 | 11 of 11 | 65 | 10% | 3m 41s | $0.190 |
| Claude Sonnet 5.5 | 11 of 11 | 17 | 10% | 4m 00s | $0.046 |
| Claude Haiku 5.5 | 8 of 11 | 55 | 13% | 2m 33s | $0.019 |
| GPT-6 Luna | 5 of 11 | 45 | 5% | 31m 20s | $0.0031 |
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.