Benchmarks · creative
MineBench
Models turn a text prompt into raw Minecraft-style block coordinates; humans vote pairwise on the rendered builds. Value is the conservative rankScore (rating − 2·RD), not raw Elo.
77 models · 77 results · data as of 2026-11-06 · minebench.ai ↗
Leaderboard
GPT-6 Astra Pro max
More from this benchmark
6 published measuresScore vs any published measure6 measures: sub-scores, splits, counts
62 more models under All models in settings.
Breakdown
darker is better within each column · click a column to sort| Model | ||||
|---|---|---|---|---|
| GPT-6 Astra Pro max | 4,198 | 885 | 416 | 6,884 |
| Gemini 4 Argon max | 1,011 | 398 | 104 | 2,029 |
| GPT-6.1 Sol Pro | 1,972 | 977 | 331 | 4,045 |
| Opus 5.5 max | 2,243 | 1,135 | 290 | 4,526 |
| GPT-6 Sol Pro max | 2,030 | 1,115 | 226 | 4,211 |
| Opus 5 max | 4,558 | 3,111 | 662 | 11,743 |
| Sonnet 5.5 max | 1,951 | 1,046 | 172 | 3,911 |
| GPT-5.6 Sol Pro max | 4,684 | 3,199 | 746 | 12,027 |
| Kimi K3.1 max | 570 | 166 | 39 | 1,163 |
| Fable 5.1 max | 2,790 | 1,803 | 299 | 6,271 |
| GPT-5.5 Pro xhigh | 7,346 | 4,144 | 1,046 | 17,756 |
| Fable 5 max | 7,218 | 3,662 | 672 | 16,045 |
| Muse Wonder (Max) | 512 | 272 | 26 | 1,073 |
| GPT-5.5 xhigh | 6,821 | 5,367 | 1,016 | 18,545 |
| Grok 4.7 max | 469 | 203 | 12 | 1,036 |
| DeepSeek V4.1 Flash max | 2,019 | 908 | 129 | 4,136 |
| Grok 4.6 max | 1,760 | 1,492 | 176 | 4,660 |
| Gemini 3.8 Flash max | 2,085 | 935 | 133 | 3,986 |
| GPT-5.4 Pro xhigh | 9,742 | 4,740 | 890 | 22,154 |
| Opus 4.8 max | 5,822 | 3,246 | 457 | 13,438 |
| GLM 5.3 Flash max | 1,882 | 896 | 113 | 3,913 |
| Gemini 3.7 Flash max | 1,657 | 973 | 145 | 3,921 |
| GLM 5.3 max | 2,230 | 1,006 | 132 | 5,080 |
| Muse Spark 1.3 max | 2,180 | 930 | 103 | 4,385 |
| GPT-6 Luna Pro max | 1,302 | 728 | 66 | 2,561 |
All results
77 rows| # | Model | Effort | MineBench |
|---|---|---|---|
| 1 | GPT-6 Astra Pro | max | 2238 ±16 |
| 2 | Gemini 4 Argon | max | 2213 ±24 |
| 3 | GPT-6.1 Sol Pro | 2178 ±18 | |
| 4 | Opus 5.5 | max | 2145 ±17 |
| 5 | GPT-6 Sol Pro | max | 2076 ±16 |
| 6 | Opus 5 | max | 2015 ±11 |
| 7 | Sonnet 5.5 | max | 2006 ±16 |
| 8 | GPT-5.6 Sol Pro | max | 2001 ±11 |
| 9 | Kimi K3.1 | max | 1999 ±32 |
| 10 | Fable 5.1 | max | 1935 ±12 |
| 11 | GPT-5.5 Pro | xhigh | 1929 ±9 |
| 12 | Fable 5 | max | 1876 ±9 |
| 13 | Muse Wonder (Max) | max | 1868 ±30 |
| 14 | GPT-5.5 | xhigh | 1853 ±8 |
| 15 | Grok 4.7 | max | 1852 ±33 |
| 16 | DeepSeek V4.1 Flash | max | 1829 ±16 |
| 17 | Grok 4.6 | max | 1818 ±14 |
| 18 | Gemini 3.8 Flash | max | 1814 ±16 |
| 19 | GPT-5.4 Pro | xhigh | 1801 ±8 |
| 20 | Opus 4.8 | max | 1792 ±9 |
| 21 | GLM 5.3 Flash | max | 1788 ±17 |
| 22 | Gemini 3.7 Flash | max | 1772 ±16 |
| 23 | GLM 5.3 | max | 1768 ±16 |
| 24 | Muse Spark 1.3 | max | 1753 ±17 |
| 25 | GPT-6 Luna Pro | max | 1752 ±19 |
| 26 | Kimi K3 | max | 1720 ±15 |
| 27 | GPT-5.6 Luna Pro | 1711 ±18 | |
| 28 | Gemini 3.1 Pro | high | 1683 ±8 |
| 29 | GPT-5.4 | xhigh | 1673 ±8 |
| 30 | Gemini 3.5 Flash | xhigh | 1659 ±10 |