Benchmark Prompts / P10
Island scene: 12 AI models, one prompt
Every Season 3 competitor received this prompt verbatim, at default settings, with one generation and no manual cleanup. Nothing below is cherry-picked — the failures ship too, because a benchmark that hides them isn't one.
12Outputs612 ballots · 16 both-bad
A low-poly island scene: small rocky island with two palm trees and a sandy beach.
Every output for this prompt
Click a tile to load the model in 3D, then switch to Wire or X-ray to read the topology. Vote on these blind →
Claude Opus 5.5Island scene Grok 4.7Island scene GPT-6 SolIsland scene GPT-6 LunaIsland scene GPT-6.1 SolIsland scene 3D-AgentIsland scene GPT-6 AstraIsland scene Claude Fable 5.1Island scene Claude Opus 5Island scene GPT-5.6 SolIsland scene Grok 4.6Island scene Claude Fable 5Island scene
Season 3 blind votes for P10
612 ballots · 16 “both bad”
| Competitor | Record | Wins | Losses |
|---|---|---|---|
| Claude Opus 5.5 | 90 | 7 | |
| 3D-Agent | 80 | 25 | |
| GPT-6 Astra | 72 | 32 | |
| Grok 4.6 | 67 | 36 | |
| GPT-6.1 Sol | 57 | 7 | |
| GPT-5.6 Sol | 51 | 53 | |
| Claude Fable 5.1 | 39 | 69 | |
| GPT-6 Sol | 39 | 51 | |
| GPT-6 Luna | 27 | 74 | |
| Claude Opus 5 | 24 | 65 | |
| Grok 4.7 | 21 | 75 | |
| Claude Fable 5 | 12 | 85 |