Benchmark Prompts / P12
Chess set: 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.
12Outputs604 ballots · 10 both-bad
A chess set arrangement: checkered board with the two front rows of white pieces placed correctly.
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 5.5Chess set Grok 4.7Chess set GPT-6 SolChess set GPT-6 LunaChess set GPT-6.1 SolChess set 3D-AgentChess set GPT-6 AstraChess set Claude Fable 5.1Chess set Claude Opus 5Chess set GPT-5.6 SolChess set Grok 4.6Chess set Claude Fable 5Chess set
Season 3 blind votes for P12
604 ballots · 10 “both bad”
| Competitor | Record | Wins | Losses |
|---|---|---|---|
| 3D-Agent | 83 | 23 | |
| Claude Opus 5.5 | 78 | 11 | |
| GPT-6 Astra | 71 | 23 | |
| Claude Fable 5.1 | 54 | 41 | |
| Claude Opus 5 | 53 | 39 | |
| GPT-5.6 Sol | 50 | 39 | |
| GPT-6 Sol | 43 | 50 | |
| GPT-6 Luna | 36 | 55 | |
| Grok 4.6 | 32 | 44 | |
| Claude Fable 5 | 19 | 80 | |
| GPT-6.1 Sol | 15 | 48 | |
| Grok 4.7 | 11 | 92 |