Mystery game puzzles — Epoch AI
Mystery game puzzles — Epoch AI is run by Epoch AI. It has ranked 142 models, of which the catalogue holds 76, scoring from 2 to 84.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 1stof 142 | GPT 6 Astra | mean_score (max) | 84 |
| 2ndof 142 | GPT 6.1 Sol | mean_score (max) | 80 |
| 3rdof 142 | GPT-6.1 Terra | mean_score (max) | 74 |
| 4thof 142 | Claude Opus 5.5 | mean_score (max) | 71 |
| 5thof 142 | Claude Sonnet 5.5 | mean_score (max) | 65 |
| 6thof 142 | Gemini 4 Argon | mean_score (high) | 64 |
| 7thof 142 | Claude Opus 5 | mean_score (max) | 59 |
| 8thof 142 | Claude Fable 5.1 | mean_score (max) | 58 |
| 9thof 142 | GPT-5.6 Sol | mean_score (max) | 58 |
| 10thof 142 | GPT 6 Sol | mean_score (max) | 56 |
| 11thof 142 | GPT-5.5 | mean_score (xhigh) | 56 |
| 12thof 142 | Claude Fable 5 | mean_score (max) | 52 |
| 15thof 142 | DeepSeek V4.1 Flash 1017 | mean_score (max) | 49 |
| 16thof 142 | Gemini 3.8 Flash | mean_score (high) | 47 |
| 17thof 142 | DeepSeek V4.1 Flash | mean_score (max) | 43 |
| 18thof 142 | DeepSeek V4 Pro 0813 | mean_score (max) | 43 |
| 19thof 142 | Qwen 3.8 Max | mean_score (xhigh) | 38 |
| 20thof 142 | Gemini 3.7 Flash | mean_score (high) | 37 |
| 22ndof 142 | GPT-5.4 | mean_score (xhigh) | 37 |
| 23rdof 142 | Claude Opus 4.8 | mean_score (max) | 36 |
| 24thof 142 | Claude Sonnet 5 | mean_score (max) | 35 |
| 25thof 142 | GPT-5.6 Terra | mean_score (max) | 35 |
| 26thof 142 | Grok 4.6 | mean_score (xhigh) | 34 |
| 27thof 142 | DeepSeek V4 Flash (0731) | mean_score (max) | 34 |
| 28thof 142 | Gemini 3.1 Pro Preview | mean_score (high) | 34 |
| 29thof 142 | GLM-5.3 | mean_score (max) | 33 |
| 32ndof 142 | Qwen3.7 Max | mean_score (max) | 32 |
| 33rdof 142 | Gemini 3.5 Flash | mean_score (high) | 32 |
| 35thof 142 | Claude Haiku 5.5 | mean_score (max) | 30 |
| 36thof 142 | Gemini 3.6 Flash | mean_score (high) | 30 |
| 37thof 142 | Grok 4.7 | mean_score (xhigh) | 29 |
| 38thof 142 | o3 | mean_score (high) | 29 |
| 43rdof 142 | Claude Opus 4.7 | mean_score (max) | 28 |
| 46thof 142 | Gemini 3 Flash | mean_score (low) | 26 |
| 47thof 142 | Kimi K3 | mean_score (max) | 26 |
| 48thof 142 | Muse Spark 1.3 | mean_score (max) | 25 |
| 51stof 142 | Claude Opus 4.6 | mean_score (max) | 25 |
| 53rdof 142 | GPT-5.2 | mean_score (high) | 23 |
| 55thof 142 | GPT-5 | mean_score (high) | 23 |
| 57thof 142 | Qwen3.6 35B A3B | mean_score (none) | 22 |
| 58thof 142 | Claude Opus 4.5 | mean_score | 22 |
| 59thof 142 | GPT-5.6 Luna | mean_score (max) | 21 |
| 60thof 142 | Claude Opus 4.1 | mean_score | 21 |
| 62ndof 142 | Qwen3.5-Flash | mean_score | 20 |
| 63rdof 142 | Nemotron 3 Ultra | mean_score | 20 |
| 67thof 142 | GPT-5.1 | mean_score (low) | 19 |
| 68thof 142 | GLM 5.2 | mean_score (low) | 19 |
| 69thof 142 | Gemini 3.5 Flash-Lite | mean_score (low) | 19 |
| 70thof 142 | Qwen3.5 397B A17B | mean_score (none) | 18 |
| 71stof 142 | Qwen3.6-Flash | mean_score (none) | 18 |
| 75thof 142 | Kimi K2.6 | mean_score | 18 |
| 77thof 142 | Qwen 3.7 Plus | mean_score | 17 |
| 79thof 142 | Qwen3.5-122B-A10B | mean_score (none) | 17 |
| 80thof 142 | Qwen 3.5 Plus | mean_score (none) | 17 |
| 83rdof 142 | DeepSeek V4 Pro | mean_score (none) | 17 |
| 84thof 142 | Claude Sonnet 4.5 | mean_score | 17 |
| 89thof 142 | Claude Sonnet 4.6 | mean_score (low) | 16 |
| 92ndof 142 | Qwen3.7 Flash | mean_score | 15 |
| 106thof 142 | GPT-4 (0613) | mean_score | 12 |
| 109thof 142 | Qwen3.6 Plus | mean_score (none) | 12 |
| 110thof 142 | GPT-4o-mini (2024-07-18) | mean_score | 12 |
| 112thof 142 | GPT-5.4 mini | mean_score (none) | 11 |
| 115thof 142 | GPT-5 mini | mean_score (minimal) | 10 |
| 116thof 142 | GPT-5 nano | mean_score (medium) | 9 |
| 117thof 142 | Qwen3 235B A22B Thinking 2507 | mean_score | 9 |
| 119thof 142 | GPT-5.4 nano | mean_score (none) | 9 |
| 120thof 142 | GLM 5.3 Flash | mean_score (max) | 8 |
| 122ndof 142 | MiniMax M3 | mean_score (none) | 8 |
| 124thof 142 | GPT 6 Luna | mean_score (max) | 7 |
| 126thof 142 | o3-mini | mean_score (high) | 7 |
| 127thof 142 | Qwen3.6 27B | mean_score (none) | 7 |
| 129thof 142 | GPT-4.1 mini | mean_score | 7 |
| 132ndof 142 | Inkling Small | mean_score (xhigh) | 6 |
| 133rdof 142 | o4-mini | mean_score (high) | 5 |
| 140thof 142 | GPT-3.5 Turbo | mean_score | 3 |
| 141stof 142 | GPT OSS 120B | mean_score (medium) | 2 |
Read from the board on 2026-11-01