Dialogue
Subcategory of Creative Writing. 202 models scored.
Model Leaderboard
All models ranked by their Dialogue subcategory score.
| # | Model | Dialogue | Creative Writing | Overall |
|---|---|---|---|---|
| 1 | GPT-6.1 Sol (Reasoning, Medium) | 99.03% | 92.05% | 95.88% |
| 2 | GPT-6 Sol | 98.54% | 91.22% | 90.05% |
| 3 | Qwen 3.8 27B (Reasoning, XHigh) | 98.46% | 89.16% | 92.91% |
| 4 | GPT-6 Sol (Reasoning, Medium) | 98.11% | 91.40% | 92.63% |
| 5 | GPT-5.6 Sol | 97.84% | 91.04% | 90.76% |
| 6 | GPT-5.6 Sol (Reasoning, Medium) | 97.62% | 91.03% | 95.15% |
| 7 | Muse Wonder (Reasoning, Medium) | 97.55% | 90.61% | 95.47% |
| 8 | Qwen 3.8 Max (Reasoning, XHigh) | 97.36% | 89.99% | 96.21% |
| 9 | Grok 4.7 (Reasoning, High) | 97.23% | 91.29% | 92.62% |
| 10 | GPT-5.6 Terra | 96.90% | 90.92% | 88.23% |
| 11 | GPT-6 Luna | 96.56% | 90.37% | 83.73% |
| 12 | GPT-6 Luna (Reasoning, High) | 96.28% | 90.86% | 93.95% |
| 13 | GPT-5.5 (Reasoning, Low) | 96.25% | 90.24% | 92.51% |
| 14 | Grok 4.6 (Reasoning, High) | 96.15% | 88.97% | 92.76% |
| 15 | Claude Opus 5 (Reasoning) | 95.93% | 88.64% | 91.15% |
| 16 | GPT-5.6 Luna | 95.55% | 89.69% | 85.06% |
| 17 | Qwen 3.8 Flash (Reasoning) | 95.50% | 88.58% | 94.24% |
| 18 | DeepSeek V4.1 Flash (Reasoning, High) | 95.45% | 88.96% | 94.30% |
| 19 | GPT-5.5 | 95.32% | 90.39% | 89.37% |
| 20 | Claude Sonnet 4.6 | 95.20% | 83.31% | 90.66% |
| 21 | GPT-6 Luna (Reasoning, Medium) | 95.20% | 90.20% | 92.40% |
| 22 | Claude Opus 5 | 95.16% | 88.13% | 89.97% |
| 23 | Claude Opus 5 (Reasoning, Low) | 95.12% | 88.87% | 92.06% |
| 24 | Z.AI GLM 5.3 Flash (Reasoning, Max) | 95.02% | 89.46% | 94.28% |
| 25 | Claude Sonnet 4.6 (Reasoning) | 94.92% | 83.09% | 93.64% |
| 26 | GPT-5.5 (Reasoning, Medium) | 94.90% | 90.26% | 93.72% |
| 27 | GPT-5.6 Luna (Reasoning, Medium) | 94.84% | 90.10% | 90.61% |
| 28 | GPT-5.4 | 94.79% | 90.94% | 84.31% |
| 29 | Muse Spark 1.3 (Reasoning, Medium) | 94.76% | 89.53% | 94.06% |
| 30 | Claude Opus 5.5 (Reasoning) | 94.62% | 91.10% | 96.43% |
| 31 | Muse Glimmer 30B (Reasoning, Medium) | 94.53% | 87.42% | 88.81% |
| 32 | GPT-5.6 Terra (Reasoning, Medium) | 94.42% | 90.72% | 92.49% |
| 33 | Z.AI GLM 5.3 (Reasoning, Max) | 94.42% | 89.99% | 95.20% |
| 34 | Xiaomi MIMO v2.6 Pro | 94.40% | 88.43% | 91.50% |
| 35 | Aion 3.0 | 93.98% | 84.64% | 88.78% |
| 36 | MiniMax M3 | 93.86% | 84.57% | 90.45% |
| 37 | Ember-1 (Reasoning, Max) | 93.86% | 89.28% | 95.52% |
| 38 | GPT-5.4 (Reasoning) | 93.69% | 91.17% | 93.85% |
| 39 | Z.AI GLM 5.2 (Reasoning, High) | 93.65% | 83.92% | 93.41% |
| 40 | Claude Opus 4.6 (Reasoning) | 93.56% | 84.55% | 95.06% |
| 41 | Xiaomi MIMO v2.6 Flash | 93.25% | 84.08% | 87.31% |
| 42 | Z.AI GLM 5 Turbo | 93.23% | 84.66% | 93.29% |
| 43 | Muse Spark 1.1 (Reasoning, Medium) | 92.70% | 88.57% | 93.81% |
| 44 | Aion 3.0 Mini | 92.64% | 84.60% | 83.69% |
| 45 | Kimi K2.6 | 92.60% | 85.47% | 92.57% |
| 46 | Qwen 3.5 397B A17B | 92.54% | 86.93% | 91.09% |
| 47 | Z.AI GLM 5.1 | 92.52% | 84.05% | 93.74% |
| 48 | Claude Opus 4.6 | 92.37% | 83.59% | 92.31% |
| 49 | Kimi K3 (Reasoning, High) | 92.23% | 87.43% | 92.99% |
| 50 | GPT-5.4 (Reasoning, Low) | 92.18% | 90.51% | 90.91% |
| 51 | Aion 3.5 Mini (Reasoning, High) | 92.04% | 88.32% | 92.36% |
| 52 | Grok 4.5 (Reasoning, High) | 92.01% | 87.43% | 94.12% |
| 53 | Qwen 3.5 9B | 91.95% | 84.35% | 84.05% |
| 54 | Qwen 3.6 Max Preview | 91.85% | 88.42% | 93.72% |
| 55 | Grok 4.5 (Reasoning, Low) | 91.72% | 87.45% | 90.94% |
| 56 | Muse Spark 1.2 (Reasoning, Medium) | 91.68% | 87.33% | 94.71% |
| 57 | Claude Sonnet 5.5 (Reasoning) | 91.50% | 90.33% | 92.77% |
| 58 | DeepSeek V4 Pro | 91.48% | 83.70% | 82.05% |
| 59 | GPT-5 | 91.30% | 86.87% | 91.48% |
| 60 | Claude Sonnet 5 | 91.19% | 82.58% | 87.34% |
| 61 | Z.AI GLM 5.3 Flash (Reasoning, Low) | 91.18% | 86.13% | 88.76% |
| 62 | DeepSeek V4 Flash 0731 (Reasoning, High) | 91.18% | 84.03% | 87.97% |
| 63 | Hy4 Preview (Reasoning, High) | 90.92% | 87.32% | 94.80% |
| 64 | Kimi K3 (Reasoning, Low) | 90.50% | 86.51% | 92.07% |
| 65 | Grok 4.3 (Reasoning) | 90.45% | 85.11% | 90.99% |
| 66 | GPT-5.4 Mini | 90.45% | 88.10% | 80.45% |
| 67 | DeepSeek V4 Pro 0813 (Reasoning, High) | 90.25% | 85.72% | 90.21% |
| 68 | GPT-5.4 Mini (Reasoning) | 90.04% | 88.66% | 89.82% |
| 69 | Aion 3.5 (Reasoning, High) | 89.96% | 86.16% | 91.90% |
| 70 | Claude Haiku 5.5 (Reasoning, Medium) | 89.69% | 88.53% | 84.17% |
| 71 | GPT-5.4 Mini (Reasoning, Low) | 89.66% | 87.72% | 83.57% |
| 72 | Claude Sonnet 4.5 | 89.63% | 84.19% | 87.54% |
| 73 | Qwen 3.5 35B | 89.61% | 83.51% | 87.01% |
| 74 | Claude Haiku 5.5 (Reasoning, Low) | 89.41% | 88.61% | 84.75% |
| 75 | Claude Opus 4 | 89.08% | 83.79% | 87.22% |
| 76 | Qwen 3.6 35B | 89.07% | 85.97% | 87.66% |
| 77 | Qwen 3.7 Max | 88.93% | 85.39% | 94.55% |
| 78 | MiniMax M2.5 | 88.93% | 81.21% | 86.71% |
| 79 | MiniMax M2.7 | 88.91% | 81.70% | 86.23% |
| 80 | Claude Opus 4.5 | 88.91% | 81.71% | 89.60% |
| 81 | Muse Spark 1.1 (Reasoning, Minimal) | 88.84% | 87.30% | 91.43% |
| 82 | Qwen 3.5 27B | 88.67% | 82.54% | 90.05% |
| 83 | Claude Opus 4.7 | 88.63% | 84.74% | 89.90% |
| 84 | Claude Haiku 5.5 (Adaptive) | 88.63% | 88.73% | 89.37% |
| 85 | DeepSeek V4 Flash | 88.48% | 83.42% | 82.02% |
| 86 | Z.AI GLM 5 | 88.46% | 83.63% | 89.60% |
| 87 | Claude Opus 4.7 (Reasoning) | 88.14% | 84.73% | 92.53% |
| 88 | DeepSeek V4 Flash (Reasoning) | 88.10% | 83.03% | 88.06% |
| 89 | Claude Opus 4.8 (Reasoning, Low) | 87.92% | 85.86% | 91.89% |
| 90 | GPT-5.1 | 87.44% | 87.20% | 90.73% |
| 91 | Qwen 3 235B A22B Instruct 2507 | 87.23% | 84.81% | 78.07% |
| 92 | Gemini 3.1 Pro (Preview) | 87.20% | 85.44% | 94.08% |
| 93 | Mistral Large 4.0 | 87.19% | 82.15% | 84.72% |
| 94 | Claude Sonnet 5 (Reasoning) | 87.18% | 81.31% | 90.40% |
| 95 | Grok 4.3 | 87.13% | 84.51% | 78.00% |
| 96 | Qwen 3.5 Flash | 87.13% | 83.81% | 85.66% |
| 97 | Claude Sonnet 5 (Reasoning, Low) | 87.10% | 81.43% | 90.16% |
| 98 | Qwen 3.5 122B | 86.76% | 83.02% | 90.32% |
| 99 | Qwen 3.6 Flash | 86.74% | 86.02% | 89.31% |
| 100 | Thinking Machines Inkling (Reasoning) | 86.71% | 80.19% | 90.78% |
| 101 | Claude Opus 4.8 (Reasoning) | 86.50% | 85.25% | 92.33% |
| 102 | Kimi K2.5 | 86.21% | 81.35% | 90.86% |
| 103 | ByteDance Seed 2.0 Lite | 85.96% | 82.35% | 84.27% |
| 104 | GPT-5 Mini | 85.94% | 80.48% | 91.31% |
| 105 | Thinking Machines Inkling | 85.63% | 80.27% | 73.96% |
| 106 | ByteDance Seed 1.6 | 85.62% | 78.43% | 89.59% |
| 107 | Mistral Small 4 (Reasoning) | 85.39% | 81.67% | 79.48% |
| 108 | Qwen 3.5 Plus (2026-04-20) | 85.08% | 85.18% | 89.79% |
| 109 | GPT-5.4 Nano (Reasoning) | 85.05% | 80.97% | 80.02% |
| 110 | Qwen 3.7 Flash (Reasoning) | 84.81% | 88.17% | 88.56% |
| 111 | Mistral Large 4.0 (Reasoning) | 84.72% | 80.54% | 92.14% |
| 112 | Writer: Palmyra X5 | 84.42% | 83.95% | 78.11% |
| 113 | DeepSeek V4 Pro (Reasoning) | 84.25% | 82.99% | 89.28% |
| 114 | DeepSeek V3 (2025-03-24) | 83.31% | 82.34% | 79.93% |
| 115 | Grok 4.20 (Reasoning) | 83.26% | 86.25% | 90.87% |
| 116 | GPT-5.4 Nano | 83.18% | 80.50% | 72.16% |
| 117 | GPT-5.4 Nano (Reasoning, Low) | 83.10% | 80.93% | 77.46% |
| 118 | Mistral Large 2 | 83.10% | 81.86% | 81.50% |
| 119 | Claude Sonnet 4 | 82.84% | 79.21% | 87.64% |
| 120 | Thinking Machines Inkling Small (Reasoning, High) | 82.72% | 74.85% | 87.39% |
| 121 | Mistral Small 4 | 82.49% | 81.12% | 75.23% |
| 122 | Claude Haiku 4.5 | 82.25% | 78.96% | 83.36% |
| 123 | Mistral Medium 3.1 | 81.51% | 81.70% | 76.08% |
| 124 | Gemma 4 26B (Reasoning) | 81.41% | 76.38% | 89.02% |
| 125 | ByteDance Seed 2.0 Mini | 81.36% | 80.11% | 85.69% |
| 126 | Mistral Medium 3.5 | 81.20% | 82.26% | 84.52% |
| 127 | Gemini 3.1 Flash Lite (Reasoning) | 81.09% | 76.31% | 85.91% |
| 128 | Gemini 3.5 Flash Lite (Reasoning) | 80.87% | 82.80% | 84.34% |
| 129 | GPT-5.2 | 80.71% | 80.36% | 89.45% |
| 130 | Mistral Large 3 | 80.22% | 81.21% | 84.29% |
| 131 | Gemma 4 31B (Reasoning) | 80.03% | 78.13% | 89.64% |
| 132 | ByteDance Seed 1.6 Flash | 79.96% | 81.51% | 70.92% |
| 133 | Xiaomi MIMO v2.5 Pro | 79.81% | 81.08% | 86.05% |
| 134 | Gemini 4 Argon (Reasoning, Medium) | 79.65% | 85.27% | 93.81% |
| 135 | Gemini 3.1 Flash Lite | 78.90% | 76.01% | 85.09% |
| 136 | o4 Mini | 78.87% | 82.04% | 86.56% |
| 137 | Qwen 3 32B | 78.83% | 81.30% | 79.37% |
| 138 | Gemini 3.5 Flash Lite (Reasoning, Minimal) | 78.76% | 82.74% | 84.45% |
| 139 | Gemini 3.7 Flash (Reasoning, Medium) | 78.74% | 83.06% | 94.85% |
| 140 | Mistral Medium 3.5 (Reasoning) | 78.62% | 81.67% | 85.32% |
| 141 | o4 Mini High | 78.26% | 82.72% | 88.78% |
| 142 | Qwen 3.6 27B | 78.25% | 82.81% | 88.33% |
| 143 | WizardLM 2 8x22b | 77.45% | 79.06% | 71.45% |
| 144 | Gemini 2.5 Pro | 77.40% | 81.03% | 88.44% |
| 145 | Xiaomi MIMO v2.5 | 77.40% | 79.16% | 83.95% |
| 146 | Grok 4.20 | 77.13% | 83.44% | 81.21% |
| 147 | GPT-4.1 | 76.96% | 81.24% | 86.82% |
| 148 | Aion 2.0 | 76.51% | 80.24% | 86.66% |
| 149 | Gemini 3.8 Flash (Reasoning, Medium) | 76.42% | 83.10% | 95.72% |
| 150 | Ministral 8B | 76.29% | 76.87% | 63.77% |
| 151 | Gemini 3.1 Flash Lite (Preview) | 75.91% | 75.78% | 85.41% |
| 152 | Ministral 3B | 75.29% | 75.49% | 59.25% |
| 153 | Ministral 3 14B | 74.59% | 79.11% | 70.45% |
| 154 | Ministral 3 3B | 74.49% | 75.45% | 65.02% |
| 155 | Z.AI GLM 4.6 | 74.32% | 78.86% | 87.64% |
| 156 | Gemma 4 26B | 73.77% | 75.17% | 84.89% |
| 157 | DeepSeek V3.2 | 73.33% | 79.95% | 82.22% |
| 158 | Z.AI GLM 4.5 | 73.22% | 76.56% | 84.95% |
| 159 | Laguna XS 2.1 | 72.93% | 77.28% | 75.42% |
| 160 | Ministral 3 8B | 72.81% | 77.26% | 69.98% |
| 161 | Gemini 3.6 Flash (Reasoning) | 71.94% | 81.02% | 94.90% |
| 162 | Z.AI GLM 4.7 | 71.01% | 78.89% | 87.67% |
| 163 | Llama 3.1 70B | 70.76% | 72.78% | 77.41% |
| 164 | Gemini 2.5 Flash | 70.64% | 77.57% | 80.61% |
| 165 | Gemma 4 31B | 70.38% | 75.59% | 85.23% |
| 166 | DeepSeek V3 (2024-12-26) | 69.79% | 77.88% | 82.62% |
| 167 | Z.AI GLM 4.7 Flash | 69.49% | 77.36% | 82.21% |
| 168 | Gemini 3 Flash (Preview, Reasoning) | 69.48% | 75.87% | 89.93% |
| 169 | Mistral Small 3.2 24B | 68.93% | 71.87% | 77.36% |
| 170 | Hermes 3 405B | 68.75% | 80.92% | 80.80% |
| 171 | DeepSeek V3.1 | 68.61% | 77.45% | 82.35% |
| 172 | Laguna S 2.1 | 68.01% | 74.16% | 79.99% |
| 173 | Gemini 3.5 Flash (Reasoning) | 67.65% | 79.87% | 93.35% |
| 174 | Nemotron 3 Super | 67.38% | 69.75% | 81.69% |
| 175 | Mistral NeMO | 67.00% | 76.72% | 63.80% |
| 176 | DeepSeek-V2 Chat | 66.95% | 77.20% | 84.09% |
| 177 | Gemini 2.5 Flash Lite (Reasoning) | 65.73% | 71.64% | 83.10% |
| 178 | Z.AI GLM 4.5 Air | 65.43% | 74.61% | 80.74% |
| 179 | Gemini 2.5 Flash (Reasoning) | 65.33% | 76.30% | 84.14% |
| 180 | Gemini 3 Flash (Preview) | 64.78% | 75.04% | 85.47% |
| 181 | Gemini 3.6 Flash (Reasoning, Minimal) | 64.68% | 79.25% | 86.68% |
| 182 | Gemma 3 27B | 64.44% | 78.79% | 75.70% |
| 183 | Qwen 2.5 72B | 64.34% | 75.16% | 73.17% |
| 184 | Cohere Command R+ (Aug. 2024) | 64.24% | 77.70% | 67.04% |
| 185 | Gemini 3.5 Flash (Reasoning, Minimal) | 63.97% | 78.55% | 85.88% |
| 186 | Arcee AI: Trinity Mini | 63.42% | 74.01% | 67.68% |
| 187 | GPT-4o, Aug. 6th (temp=0) | 63.20% | 73.65% | 82.18% |
| 188 | Cydonia 24B V4.1 | 62.72% | 74.19% | 72.68% |
| 189 | Gemini 2.5 Flash Lite | 62.67% | 75.05% | 79.91% |
| 190 | GPT-5 Nano | 62.32% | 67.04% | 80.16% |
| 191 | Qwen 3.5 Plus (2026-02-15) | 61.06% | 77.07% | 86.17% |
| 192 | Nemotron 3 Nano | 60.70% | 65.87% | 74.50% |
| 193 | Inception Mercury 2 | 59.94% | 68.31% | 81.99% |
| 194 | GPT-4o, Aug. 6th (temp=1) | 59.42% | 75.50% | 81.28% |
| 195 | Hermes 3 70B | 59.25% | 77.41% | 69.74% |
| 196 | GPT-4.1 Mini | 57.72% | 74.52% | 81.40% |
| 197 | GPT-4o Mini (temp=0) | 56.91% | 73.10% | 76.86% |
| 198 | GPT-OSS 120B | 55.74% | 67.85% | 84.81% |
| 199 | GPT-4.1 Nano | 55.61% | 71.81% | 69.90% |
| 200 | Gemma 3 12B | 54.87% | 75.38% | 76.07% |
| 201 | GPT-4o Mini (temp=1) | 53.50% | 74.37% | 77.82% |
| 202 | Gemma 3 4B | 52.95% | 72.10% | 66.33% |