Chess puzzles — Epoch AI
Chess puzzles — Epoch AI is run by Epoch AI. It has ranked 154 models, of which the catalogue holds 81, scoring from 0 to 100.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 2ndof 154 | GPT 6 Astra | mean_score (max) | 74 |
| 4thof 154 | GPT 6.1 Sol | mean_score (max) | 68 |
| 5thof 154 | Claude Opus 5.5 | mean_score (max) | 67 |
| 6thof 154 | Gemini 3.7 Flash | mean_score (high) | 66 |
| 7thof 154 | GPT-5.5 | mean_score (xhigh) | 63 |
| 8thof 154 | GPT-5.5 Pro | mean_score (xhigh) | 62 |
| 9thof 154 | GPT-5.2 Pro | mean_score (xhigh) | 60 |
| 10thof 154 | GPT-5.4 Pro | mean_score (xhigh) | 59 |
| 11thof 154 | Gemini 4 Argon | mean_score (high) | 59 |
| 12thof 154 | GPT-5.6 Sol | mean_score (max) | 59 |
| 13thof 154 | Claude Sonnet 5.5 | mean_score (max) | 59 |
| 14thof 154 | Gemini 3.8 Flash | mean_score (high) | 58 |
| 15thof 154 | GPT 6 Sol | mean_score (max) | 56 |
| 16thof 154 | Claude Haiku 5.5 | mean_score (max) | 55 |
| 17thof 154 | Claude Opus 4.8 | mean_score (max) | 52 |
| 18thof 154 | Gemini 3.1 Pro Preview | mean_score (high) | 51 |
| 19thof 154 | GPT-5.2 | mean_score (xhigh) | 48 |
| 20thof 154 | GPT-5 Pro | mean_score (high) | 48 |
| 21stof 154 | GPT-5.4 | mean_score (xhigh) | 48 |
| 22ndof 154 | Claude Fable 5.1 | mean_score (max) | 48 |
| 23rdof 154 | DeepSeek V4.1 Flash | mean_score (max) | 48 |
| 24thof 154 | GPT-5.6 Terra | mean_score (max) | 47 |
| 25thof 154 | Claude Fable 5 | mean_score (max) | 47 |
| 26thof 154 | DeepSeek V4.1 Flash 1017 | mean_score (max) | 46 |
| 27thof 154 | Qwen 3.8 Max | mean_score (xhigh) | 46 |
| 28thof 154 | DeepSeek V4 Pro 0813 | mean_score (max) | 46 |
| 29thof 154 | GPT-5.4 nano | mean_score (xhigh) | 45 |
| 31stof 154 | Gemini 3.5 Flash | mean_score (high) | 45 |
| 33rdof 154 | Claude Sonnet 5 | mean_score (max) | 45 |
| 34thof 154 | GPT-5.6 Luna | mean_score (max) | 43 |
| 35thof 154 | Claude Opus 5 | mean_score (max) | 43 |
| 36thof 154 | GPT-5.1 | mean_score (high) | 42 |
| 39thof 154 | Gemini 3.6 Flash | mean_score (low) | 42 |
| 40thof 154 | Claude Opus 4.7 | mean_score (max) | 41 |
| 41stof 154 | Grok 4.7 | mean_score (xhigh) | 41 |
| 43rdof 154 | Gemini 3 Flash | mean_score (high) | 39 |
| 44thof 154 | GLM-5.3 | mean_score (max) | 37 |
| 45thof 154 | Kimi K3 | mean_score (high) | 37 |
| 46thof 154 | o3 | mean_score (medium) | 37 |
| 48thof 154 | Muse Spark 1.3 | mean_score (max) | 37 |
| 53rdof 154 | GPT-5 | mean_score (high) | 36 |
| 56thof 154 | GPT 6 Luna | mean_score (max) | 36 |
| 58thof 154 | Grok 4.5 | mean_score (high) | 35 |
| 60thof 154 | Muse Spark 1.1 | mean_score (xhigh) | 34 |
| 63rdof 154 | Grok 4.6 | mean_score (xhigh) | 33 |
| 66thof 154 | DeepSeek V4 Flash (0731) | mean_score (max) | 32 |
| 71stof 154 | Grok 4.20 | mean_score | 29 |
| 76thof 154 | Muse Spark 1.2 | mean_score (xhigh) | 27 |
| 80thof 154 | Qwen3.7 Max | mean_score (max) | 26 |
| 81stof 154 | GPT-5 mini | mean_score (high) | 24 |
| 83rdof 154 | GPT-5.4 mini | mean_score (high) | 23 |
| 84thof 154 | Claude Sonnet 4.6 | mean_score (max) | 23 |
| 85thof 154 | GLM 5.2 | mean_score (max) | 22 |
| 86thof 154 | o4-mini | mean_score (medium) | 22 |
| 87thof 154 | GPT-5 nano | mean_score (high) | 22 |
| 89thof 154 | GLM-4.7 | mean_score | 21 |
| 90thof 154 | GLM 5.3 Flash | mean_score (max) | 21 |
| 91stof 154 | Claude Opus 4.6 | mean_score (max) | 21 |
| 92ndof 154 | Gemini 3.5 Flash-Lite | mean_score (low) | 21 |
| 93rdof 154 | Inkling | mean_score (xhigh) | 21 |
| 94thof 154 | GLM 5.1 | mean_score | 20 |
| 95thof 154 | Kimi K2.6 | mean_score | 20 |
| 99thof 154 | DeepSeek V4 Pro | mean_score (max) | 19 |
| 104thof 154 | Grok 4.3 | mean_score (high) | 18 |
| 110thof 154 | o3-mini | mean_score (high) | 15 |
| 114thof 154 | Kimi K2.7 Code | mean_score | 14 |
| 115thof 154 | Claude Opus 4.5 | mean_score | 14 |
| 116thof 154 | Qwen3.5-Flash | mean_score | 14 |
| 117thof 154 | Qwen3.6 Plus | mean_score | 14 |
| 119thof 154 | MiniMax M3 | mean_score | 13 |
| 120thof 154 | Qwen3.6-Flash | mean_score | 12 |
| 122ndof 154 | o1 | mean_score (high) | 10 |
| 124thof 154 | Claude Sonnet 4.5 | mean_score | 10 |
| 129thof 154 | Claude Haiku 4.5 | mean_score | 8 |
| 130thof 154 | Qwen 3.5 Plus | mean_score | 8 |
| 134thof 154 | GPT-4o (2024-08-06) | mean_score | 6 |
| 143rdof 154 | GPT-4o-mini (2024-07-18) | mean_score | 3 |
| 144thof 154 | GPT-4.1 fine-tuned | mean_score | 3 |
| 147thof 154 | Claude 3 Opus | mean_score | 3 |
| 152ndof 154 | GPT-4.1 mini | mean_score | 100 |
| 153rdof 154 | GPT-4.1 nano | mean_score | 0 |
Read from the board on 2026-10-31