Purple Prose
Subcategory of Creative Writing. 202 models scored.
Model Leaderboard
All models ranked by their Purple Prose subcategory score.
| # | Model | Purple Prose | Creative Writing | Overall |
|---|---|---|---|---|
| 1 | GPT-6 Sol (Reasoning, Medium) | 99.52% | 91.40% | 92.63% |
| 2 | GPT-6 Sol | 99.42% | 91.22% | 90.05% |
| 3 | GPT-6.1 Sol (Reasoning, Medium) | 99.32% | 92.05% | 95.88% |
| 4 | GPT-5.6 Sol (Reasoning, Medium) | 99.20% | 91.03% | 95.15% |
| 5 | GPT-6 Luna (Reasoning, High) | 99.04% | 90.86% | 93.95% |
| 6 | GPT-5.6 Luna (Reasoning, Medium) | 98.87% | 90.10% | 90.61% |
| 7 | GPT-5.6 Sol | 98.82% | 91.04% | 90.76% |
| 8 | GPT-5.6 Terra | 98.78% | 90.92% | 88.23% |
| 9 | GPT-5.6 Terra (Reasoning, Medium) | 98.50% | 90.72% | 92.49% |
| 10 | GPT-6 Luna (Reasoning, Medium) | 98.46% | 90.20% | 92.40% |
| 11 | Z.AI GLM 5.3 (Reasoning, Max) | 98.40% | 89.99% | 95.20% |
| 12 | GPT-5.5 | 98.39% | 90.39% | 89.37% |
| 13 | GPT-6 Luna | 98.33% | 90.37% | 83.73% |
| 14 | GPT-5.6 Luna | 98.25% | 89.69% | 85.06% |
| 15 | GPT-5.5 (Reasoning, Low) | 98.18% | 90.24% | 92.51% |
| 16 | GPT-5.5 (Reasoning, Medium) | 98.18% | 90.26% | 93.72% |
| 17 | Muse Wonder (Reasoning, Medium) | 98.13% | 90.61% | 95.47% |
| 18 | Qwen 3.6 Max Preview | 98.07% | 88.42% | 93.72% |
| 19 | Qwen 3.7 Flash (Reasoning) | 98.07% | 88.17% | 88.56% |
| 20 | Qwen 3.5 Plus (2026-04-20) | 97.91% | 85.18% | 89.79% |
| 21 | Qwen 3.8 27B (Reasoning, XHigh) | 97.88% | 89.16% | 92.91% |
| 22 | Muse Spark 1.3 (Reasoning, Medium) | 97.83% | 89.53% | 94.06% |
| 23 | Grok 4.7 (Reasoning, High) | 97.64% | 91.29% | 92.62% |
| 24 | Qwen 3.8 Max (Reasoning, XHigh) | 97.63% | 89.99% | 96.21% |
| 25 | Ember-1 (Reasoning, Max) | 97.57% | 89.28% | 95.52% |
| 26 | o4 Mini High | 97.56% | 82.72% | 88.78% |
| 27 | GPT-5.4 (Reasoning) | 97.55% | 91.17% | 93.85% |
| 28 | DeepSeek V4.1 Flash (Reasoning, High) | 97.41% | 88.96% | 94.30% |
| 29 | o4 Mini | 97.30% | 82.04% | 86.56% |
| 30 | Z.AI GLM 5.3 Flash (Reasoning, Max) | 97.22% | 89.46% | 94.28% |
| 31 | Qwen 3.6 Flash | 96.80% | 86.02% | 89.31% |
| 32 | Qwen 3.6 27B | 96.75% | 82.81% | 88.33% |
| 33 | GPT-5.4 | 96.74% | 90.94% | 84.31% |
| 34 | GPT-5.4 (Reasoning, Low) | 96.68% | 90.51% | 90.91% |
| 35 | Qwen 3.5 9B | 96.66% | 84.35% | 84.05% |
| 36 | Qwen 3.6 35B | 96.65% | 85.97% | 87.66% |
| 37 | Kimi K3 (Reasoning, High) | 96.61% | 87.43% | 92.99% |
| 38 | Xiaomi MIMO v2.6 Pro | 96.55% | 88.43% | 91.50% |
| 39 | Muse Glimmer 30B (Reasoning, Medium) | 96.53% | 87.42% | 88.81% |
| 40 | Claude Opus 5 (Reasoning, Low) | 96.48% | 88.87% | 92.06% |
| 41 | Aion 3.5 Mini (Reasoning, High) | 96.04% | 88.32% | 92.36% |
| 42 | Aion 3.5 (Reasoning, High) | 96.01% | 86.16% | 91.90% |
| 43 | Muse Spark 1.2 (Reasoning, Medium) | 95.96% | 87.33% | 94.71% |
| 44 | Qwen 3.8 Flash (Reasoning) | 95.83% | 88.58% | 94.24% |
| 45 | Gemini 3.6 Flash (Reasoning) | 95.75% | 81.02% | 94.90% |
| 46 | Claude Sonnet 5.5 (Reasoning) | 95.75% | 90.33% | 92.77% |
| 47 | Qwen 3 235B A22B Instruct 2507 | 95.70% | 84.81% | 78.07% |
| 48 | Gemini 3.7 Flash (Reasoning, Medium) | 95.63% | 83.06% | 94.85% |
| 49 | Claude Opus 5 (Reasoning) | 95.60% | 88.64% | 91.15% |
| 50 | Hy4 Preview (Reasoning, High) | 95.57% | 87.32% | 94.80% |
| 51 | Kimi K2.6 | 95.55% | 85.47% | 92.57% |
| 52 | Muse Spark 1.1 (Reasoning, Medium) | 95.54% | 88.57% | 93.81% |
| 53 | Grok 4.3 | 95.49% | 84.51% | 78.00% |
| 54 | Qwen 3.7 Max | 95.46% | 85.39% | 94.55% |
| 55 | DeepSeek V3 (2025-03-24) | 95.33% | 82.34% | 79.93% |
| 56 | Claude Opus 5.5 (Reasoning) | 95.22% | 91.10% | 96.43% |
| 57 | Xiaomi MIMO v2.6 Flash | 95.19% | 84.08% | 87.31% |
| 58 | GPT-5.4 Mini (Reasoning, Low) | 95.16% | 87.72% | 83.57% |
| 59 | Claude Opus 5 | 95.11% | 88.13% | 89.97% |
| 60 | Gemini 3.8 Flash (Reasoning, Medium) | 95.11% | 83.10% | 95.72% |
| 61 | ByteDance Seed 1.6 Flash | 95.11% | 81.51% | 70.92% |
| 62 | Gemini 3.1 Pro (Preview) | 95.06% | 85.44% | 94.08% |
| 63 | DeepSeek V4 Pro 0813 (Reasoning, High) | 95.04% | 85.72% | 90.21% |
| 64 | Kimi K3 (Reasoning, Low) | 95.04% | 86.51% | 92.07% |
| 65 | Grok 4.5 (Reasoning, Low) | 95.00% | 87.45% | 90.94% |
| 66 | Writer: Palmyra X5 | 94.98% | 83.95% | 78.11% |
| 67 | Muse Spark 1.1 (Reasoning, Minimal) | 94.91% | 87.30% | 91.43% |
| 68 | Qwen 3.5 397B A17B | 94.84% | 86.93% | 91.09% |
| 69 | Gemini 3.6 Flash (Reasoning, Minimal) | 94.74% | 79.25% | 86.68% |
| 70 | Qwen 3.5 Flash | 94.71% | 83.81% | 85.66% |
| 71 | Grok 4.6 (Reasoning, High) | 94.70% | 88.97% | 92.76% |
| 72 | GPT-5.4 Mini | 94.65% | 88.10% | 80.45% |
| 73 | Mistral Medium 3.1 | 94.59% | 81.70% | 76.08% |
| 74 | Qwen 3.5 27B | 94.46% | 82.54% | 90.05% |
| 75 | Claude Haiku 5.5 (Adaptive) | 94.28% | 88.73% | 89.37% |
| 76 | Qwen 3.5 122B | 94.27% | 83.02% | 90.32% |
| 77 | GPT-5.4 Mini (Reasoning) | 94.24% | 88.66% | 89.82% |
| 78 | Grok 4.20 (Reasoning) | 94.23% | 86.25% | 90.87% |
| 79 | Claude Opus 4.8 (Reasoning, Low) | 94.06% | 85.86% | 91.89% |
| 80 | Grok 4.3 (Reasoning) | 94.05% | 85.11% | 90.99% |
| 81 | Gemini 4 Argon (Reasoning, Medium) | 94.01% | 85.27% | 93.81% |
| 82 | Claude Opus 4.8 (Reasoning) | 93.97% | 85.25% | 92.33% |
| 83 | Grok 4.5 (Reasoning, High) | 93.96% | 87.43% | 94.12% |
| 84 | Mistral Small 4 | 93.71% | 81.12% | 75.23% |
| 85 | Qwen 3 32B | 93.62% | 81.30% | 79.37% |
| 86 | Hermes 3 405B | 93.61% | 80.92% | 80.80% |
| 87 | Qwen 3.5 35B | 93.58% | 83.51% | 87.01% |
| 88 | GPT-5.1 | 93.57% | 87.20% | 90.73% |
| 89 | GPT-4.1 | 93.57% | 81.24% | 86.82% |
| 90 | DeepSeek V4 Flash 0731 (Reasoning, High) | 93.33% | 84.03% | 87.97% |
| 91 | Mistral Medium 3.5 | 93.31% | 82.26% | 84.52% |
| 92 | Kimi K2.5 | 93.06% | 81.35% | 90.86% |
| 93 | Thinking Machines Inkling | 92.99% | 80.27% | 73.96% |
| 94 | Claude Opus 4.7 (Reasoning) | 92.86% | 84.73% | 92.53% |
| 95 | Z.AI GLM 5.3 Flash (Reasoning, Low) | 92.77% | 86.13% | 88.76% |
| 96 | Ministral 3 14B | 92.74% | 79.11% | 70.45% |
| 97 | Mistral Medium 3.5 (Reasoning) | 92.66% | 81.67% | 85.32% |
| 98 | Mistral Large 3 | 92.61% | 81.21% | 84.29% |
| 99 | Xiaomi MIMO v2.5 Pro | 92.48% | 81.08% | 86.05% |
| 100 | Gemini 3.5 Flash Lite (Reasoning, Minimal) | 92.36% | 82.74% | 84.45% |
| 101 | Mistral Large 4.0 | 92.33% | 82.15% | 84.72% |
| 102 | Claude Haiku 5.5 (Reasoning, Medium) | 92.30% | 88.53% | 84.17% |
| 103 | DeepSeek V4 Flash (Reasoning) | 92.27% | 83.03% | 88.06% |
| 104 | Aion 3.0 | 92.17% | 84.64% | 88.78% |
| 105 | GPT-5 | 92.06% | 86.87% | 91.48% |
| 106 | Mistral Large 2 | 91.94% | 81.86% | 81.50% |
| 107 | Claude Opus 4.6 (Reasoning) | 91.88% | 84.55% | 95.06% |
| 108 | DeepSeek V4 Pro | 91.83% | 83.70% | 82.05% |
| 109 | Gemini 3.5 Flash Lite (Reasoning) | 91.78% | 82.80% | 84.34% |
| 110 | Gemini 3.5 Flash (Reasoning) | 91.71% | 79.87% | 93.35% |
| 111 | DeepSeek V4 Pro (Reasoning) | 91.70% | 82.99% | 89.28% |
| 112 | Claude Haiku 5.5 (Reasoning, Low) | 91.66% | 88.61% | 84.75% |
| 113 | DeepSeek-V2 Chat | 91.52% | 77.20% | 84.09% |
| 114 | Claude Opus 4.7 | 91.45% | 84.74% | 89.90% |
| 115 | DeepSeek V3 (2024-12-26) | 91.42% | 77.88% | 82.62% |
| 116 | Mistral Large 4.0 (Reasoning) | 91.39% | 80.54% | 92.14% |
| 117 | Z.AI GLM 4.7 | 91.38% | 78.89% | 87.67% |
| 118 | GPT-4o Mini (temp=1) | 91.21% | 74.37% | 77.82% |
| 119 | Claude Opus 4.6 | 91.07% | 83.59% | 92.31% |
| 120 | MiniMax M3 | 90.97% | 84.57% | 90.45% |
| 121 | DeepSeek V4 Flash | 90.79% | 83.42% | 82.02% |
| 122 | ByteDance Seed 2.0 Lite | 90.78% | 82.35% | 84.27% |
| 123 | Ministral 3 8B | 90.74% | 77.26% | 69.98% |
| 124 | Aion 3.0 Mini | 90.67% | 84.60% | 83.69% |
| 125 | DeepSeek V3.2 | 90.65% | 79.95% | 82.22% |
| 126 | Mistral Small 4 (Reasoning) | 90.49% | 81.67% | 79.48% |
| 127 | Z.AI GLM 5.2 (Reasoning, High) | 90.45% | 83.92% | 93.41% |
| 128 | Claude Opus 4 | 90.42% | 83.79% | 87.22% |
| 129 | Grok 4.20 | 90.27% | 83.44% | 81.21% |
| 130 | Ministral 3B | 90.17% | 75.49% | 59.25% |
| 131 | Claude Sonnet 4.6 | 90.14% | 83.31% | 90.66% |
| 132 | Gemini 2.5 Pro | 90.00% | 81.03% | 88.44% |
| 133 | Mistral NeMO | 89.84% | 76.72% | 63.80% |
| 134 | Aion 2.0 | 89.80% | 80.24% | 86.66% |
| 135 | Z.AI GLM 5.1 | 89.65% | 84.05% | 93.74% |
| 136 | Z.AI GLM 4.7 Flash | 89.60% | 77.36% | 82.21% |
| 137 | Ministral 3 3B | 89.60% | 75.45% | 65.02% |
| 138 | Thinking Machines Inkling (Reasoning) | 89.46% | 80.19% | 90.78% |
| 139 | Claude Sonnet 4.5 | 89.39% | 84.19% | 87.54% |
| 140 | Ministral 8B | 89.29% | 76.87% | 63.77% |
| 141 | Claude Sonnet 4.6 (Reasoning) | 89.19% | 83.09% | 93.64% |
| 142 | Gemini 3.5 Flash (Reasoning, Minimal) | 88.91% | 78.55% | 85.88% |
| 143 | GPT-4o Mini (temp=0) | 88.83% | 73.10% | 76.86% |
| 144 | Xiaomi MIMO v2.5 | 88.79% | 79.16% | 83.95% |
| 145 | Z.AI GLM 5 | 88.49% | 83.63% | 89.60% |
| 146 | Gemma 3 27B | 88.37% | 78.79% | 75.70% |
| 147 | GPT-4o, Aug. 6th (temp=1) | 87.96% | 75.50% | 81.28% |
| 148 | GPT-4.1 Mini | 87.91% | 74.52% | 81.40% |
| 149 | Z.AI GLM 5 Turbo | 87.80% | 84.66% | 93.29% |
| 150 | ByteDance Seed 2.0 Mini | 87.57% | 80.11% | 85.69% |
| 151 | GPT-5.2 | 87.55% | 80.36% | 89.45% |
| 152 | Qwen 3.5 Plus (2026-02-15) | 87.42% | 77.07% | 86.17% |
| 153 | Cohere Command R+ (Aug. 2024) | 87.40% | 77.70% | 67.04% |
| 154 | Hermes 3 70B | 87.04% | 77.41% | 69.74% |
| 155 | ByteDance Seed 1.6 | 87.01% | 78.43% | 89.59% |
| 156 | Gemma 3 12B | 86.76% | 75.38% | 76.07% |
| 157 | Laguna XS 2.1 | 86.56% | 77.28% | 75.42% |
| 158 | DeepSeek V3.1 | 86.19% | 77.45% | 82.35% |
| 159 | Arcee AI: Trinity Mini | 85.69% | 74.01% | 67.68% |
| 160 | WizardLM 2 8x22b | 85.19% | 79.06% | 71.45% |
| 161 | Gemma 4 31B (Reasoning) | 85.05% | 78.13% | 89.64% |
| 162 | GPT-4o, Aug. 6th (temp=0) | 84.90% | 73.65% | 82.18% |
| 163 | GPT-5.4 Nano (Reasoning, Low) | 84.72% | 80.93% | 77.46% |
| 164 | MiniMax M2.5 | 84.71% | 81.21% | 86.71% |
| 165 | Mistral Small 3.2 24B | 84.70% | 71.87% | 77.36% |
| 166 | Qwen 2.5 72B | 84.37% | 75.16% | 73.17% |
| 167 | Z.AI GLM 4.6 | 84.37% | 78.86% | 87.64% |
| 168 | Gemini 3 Flash (Preview) | 84.15% | 75.04% | 85.47% |
| 169 | GPT-5.4 Nano (Reasoning) | 84.09% | 80.97% | 80.02% |
| 170 | MiniMax M2.7 | 83.74% | 81.70% | 86.23% |
| 171 | GPT-5.4 Nano | 83.70% | 80.50% | 72.16% |
| 172 | Gemini 2.5 Flash | 83.37% | 77.57% | 80.61% |
| 173 | Gemma 4 31B | 83.35% | 75.59% | 85.23% |
| 174 | Claude Opus 4.5 | 83.27% | 81.71% | 89.60% |
| 175 | GPT-5 Mini | 83.16% | 80.48% | 91.31% |
| 176 | Gemma 3 4B | 83.05% | 72.10% | 66.33% |
| 177 | GPT-4.1 Nano | 82.65% | 71.81% | 69.90% |
| 178 | Claude Sonnet 5 (Reasoning) | 82.55% | 81.31% | 90.40% |
| 179 | Gemini 3.1 Flash Lite | 82.35% | 76.01% | 85.09% |
| 180 | Claude Sonnet 5 (Reasoning, Low) | 82.23% | 81.43% | 90.16% |
| 181 | Gemini 3.1 Flash Lite (Preview) | 82.09% | 75.78% | 85.41% |
| 182 | Claude Haiku 4.5 | 81.79% | 78.96% | 83.36% |
| 183 | Z.AI GLM 4.5 Air | 81.79% | 74.61% | 80.74% |
| 184 | Z.AI GLM 4.5 | 81.73% | 76.56% | 84.95% |
| 185 | Gemini 2.5 Flash (Reasoning) | 81.71% | 76.30% | 84.14% |
| 186 | Laguna S 2.1 | 81.48% | 74.16% | 79.99% |
| 187 | Claude Sonnet 5 | 81.40% | 82.58% | 87.34% |
| 188 | Thinking Machines Inkling Small (Reasoning, High) | 81.02% | 74.85% | 87.39% |
| 189 | Gemini 3.1 Flash Lite (Reasoning) | 80.96% | 76.31% | 85.91% |
| 190 | Gemini 2.5 Flash Lite | 79.16% | 75.05% | 79.91% |
| 191 | Gemini 3 Flash (Preview, Reasoning) | 78.95% | 75.87% | 89.93% |
| 192 | Claude Sonnet 4 | 78.57% | 79.21% | 87.64% |
| 193 | Gemma 4 26B (Reasoning) | 78.14% | 76.38% | 89.02% |
| 194 | Gemma 4 26B | 77.99% | 75.17% | 84.89% |
| 195 | Cydonia 24B V4.1 | 77.12% | 74.19% | 72.68% |
| 196 | Gemini 2.5 Flash Lite (Reasoning) | 72.59% | 71.64% | 83.10% |
| 197 | Llama 3.1 70B | 71.74% | 72.78% | 77.41% |
| 198 | Nemotron 3 Super | 70.69% | 69.75% | 81.69% |
| 199 | GPT-OSS 120B | 70.06% | 67.85% | 84.81% |
| 200 | Inception Mercury 2 | 67.97% | 68.31% | 81.99% |
| 201 | GPT-5 Nano | 66.77% | 67.04% | 80.16% |
| 202 | Nemotron 3 Nano | 59.84% | 65.87% | 74.50% |