Creative Writing
18 scenarios across 6 subcategories. 202 models scored.
Subcategories
| Subcategory | Avg Score | Best Model | Best Score |
|---|---|---|---|
| AI-isms | 84.69% | GPT-6 Sol | 95.76% |
| Prose Variety | 69.23% | Writer: Palmyra X5 | 83.01% |
| Dialogue | 82.78% | GPT-6.1 Sol (Reasoning, Medium) | 99.03% |
| Purple Prose | 90.73% | GPT-6 Sol (Reasoning, Medium) | 99.52% |
| Mechanical Style | 87.72% | GPT-5.6 Sol (Reasoning, Medium) | 99.81% |
| Clichés | 79.92% | Claude Opus 5.5 (Reasoning) | 93.17% |
Model Leaderboard
All models ranked by their Creative Writing category score.
| # | Model | Creative Writing | AI-isms | Prose Variety | Dialogue | Purple Prose | Mechanical Style | Clichés | Overall |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GPT-6.1 Sol (Reasoning, Medium) | 92.05% | 95.35% | 73.51% | 99.03% | 99.32% | 99.61% | 85.49% | 95.88% |
| 2 | GPT-6 Sol (Reasoning, Medium) | 91.40% | 95.74% | 71.77% | 98.11% | 99.52% | 99.13% | 84.11% | 92.63% |
| 3 | Grok 4.7 (Reasoning, High) | 91.29% | 95.30% | 68.10% | 97.23% | 97.64% | 97.19% | 92.27% | 92.62% |
| 4 | GPT-6 Sol | 91.22% | 95.76% | 70.80% | 98.54% | 99.42% | 99.52% | 83.26% | 90.05% |
| 5 | GPT-5.4 (Reasoning) | 91.17% | 91.56% | 79.74% | 93.69% | 97.55% | 97.97% | 86.52% | 93.85% |
| 6 | Claude Opus 5.5 (Reasoning) | 91.10% | 92.05% | 74.26% | 94.62% | 95.22% | 97.29% | 93.17% | 96.43% |
| 7 | GPT-5.6 Sol | 91.04% | 92.68% | 72.39% | 97.84% | 98.82% | 99.63% | 84.87% | 90.76% |
| 8 | GPT-5.6 Sol (Reasoning, Medium) | 91.03% | 92.56% | 73.84% | 97.62% | 99.20% | 99.81% | 83.12% | 95.15% |
| 9 | GPT-5.4 | 90.94% | 91.06% | 81.26% | 94.79% | 96.74% | 97.49% | 84.30% | 84.31% |
| 10 | GPT-5.6 Terra | 90.92% | 92.94% | 73.87% | 96.90% | 98.78% | 99.54% | 83.46% | 88.23% |
| 11 | GPT-6 Luna (Reasoning, High) | 90.86% | 93.89% | 70.61% | 96.28% | 99.04% | 99.31% | 86.04% | 93.95% |
| 12 | GPT-5.6 Terra (Reasoning, Medium) | 90.72% | 92.93% | 74.76% | 94.42% | 98.50% | 99.04% | 84.66% | 92.49% |
| 13 | Muse Wonder (Reasoning, Medium) | 90.61% | 93.29% | 69.00% | 97.55% | 98.13% | 97.82% | 87.87% | 95.47% |
| 14 | GPT-5.4 (Reasoning, Low) | 90.51% | 91.02% | 80.88% | 92.18% | 96.68% | 97.75% | 84.51% | 90.91% |
| 15 | GPT-5.5 | 90.39% | 91.50% | 76.57% | 95.32% | 98.39% | 99.17% | 81.40% | 89.37% |
| 16 | GPT-6 Luna | 90.37% | 93.43% | 70.16% | 96.56% | 98.33% | 99.17% | 84.57% | 83.73% |
| 17 | Claude Sonnet 5.5 (Reasoning) | 90.33% | 91.78% | 72.73% | 91.50% | 95.75% | 97.27% | 92.94% | 92.77% |
| 18 | GPT-5.5 (Reasoning, Medium) | 90.26% | 92.18% | 75.69% | 94.90% | 98.18% | 99.27% | 81.35% | 93.72% |
| 19 | GPT-5.5 (Reasoning, Low) | 90.24% | 91.49% | 77.34% | 96.25% | 98.18% | 99.26% | 78.89% | 92.51% |
| 20 | GPT-6 Luna (Reasoning, Medium) | 90.20% | 93.65% | 70.40% | 95.20% | 98.46% | 98.76% | 84.75% | 92.40% |
| 21 | GPT-5.6 Luna (Reasoning, Medium) | 90.10% | 92.47% | 72.24% | 94.84% | 98.87% | 99.54% | 82.62% | 90.61% |
| 22 | Qwen 3.8 Max (Reasoning, XHigh) | 89.99% | 92.92% | 67.30% | 97.36% | 97.63% | 99.36% | 85.37% | 96.21% |
| 23 | Z.AI GLM 5.3 (Reasoning, Max) | 89.99% | 93.36% | 76.16% | 94.42% | 98.40% | 84.48% | 93.12% | 95.20% |
| 24 | GPT-5.6 Luna | 89.69% | 91.99% | 71.83% | 95.55% | 98.25% | 99.39% | 81.14% | 85.06% |
| 25 | Muse Spark 1.3 (Reasoning, Medium) | 89.53% | 91.37% | 69.28% | 94.76% | 97.83% | 96.56% | 87.39% | 94.06% |
| 26 | Z.AI GLM 5.3 Flash (Reasoning, Max) | 89.46% | 93.80% | 74.99% | 95.02% | 97.22% | 85.62% | 90.12% | 94.28% |
| 27 | Ember-1 (Reasoning, Max) | 89.28% | 93.41% | 76.75% | 93.86% | 97.57% | 85.67% | 88.42% | 95.52% |
| 28 | Qwen 3.8 27B (Reasoning, XHigh) | 89.16% | 94.22% | 62.15% | 98.46% | 97.88% | 98.46% | 83.75% | 92.91% |
| 29 | Grok 4.6 (Reasoning, High) | 88.97% | 92.81% | 67.01% | 96.15% | 94.70% | 95.88% | 87.29% | 92.76% |
| 30 | DeepSeek V4.1 Flash (Reasoning, High) | 88.96% | 93.14% | 74.11% | 95.45% | 97.41% | 87.78% | 85.88% | 94.30% |
| 31 | Claude Opus 5 (Reasoning, Low) | 88.87% | 93.69% | 75.41% | 95.12% | 96.48% | 84.59% | 87.91% | 92.06% |
| 32 | Claude Haiku 5.5 (Adaptive) | 88.73% | 90.76% | 74.30% | 88.63% | 94.28% | 95.50% | 88.92% | 89.37% |
| 33 | GPT-5.4 Mini (Reasoning) | 88.66% | 90.82% | 74.97% | 90.04% | 94.24% | 98.70% | 83.22% | 89.82% |
| 34 | Claude Opus 5 (Reasoning) | 88.64% | 93.54% | 75.80% | 95.93% | 95.60% | 83.74% | 87.25% | 91.15% |
| 35 | Claude Haiku 5.5 (Reasoning, Low) | 88.61% | 91.04% | 75.59% | 89.41% | 91.66% | 94.70% | 89.27% | 84.75% |
| 36 | Qwen 3.8 Flash (Reasoning) | 88.58% | 93.16% | 65.55% | 95.50% | 95.83% | 97.71% | 83.71% | 94.24% |
| 37 | Muse Spark 1.1 (Reasoning, Medium) | 88.57% | 91.62% | 72.77% | 92.70% | 95.54% | 93.88% | 84.93% | 93.81% |
| 38 | Claude Haiku 5.5 (Reasoning, Medium) | 88.53% | 90.98% | 74.86% | 89.69% | 92.30% | 94.95% | 88.38% | 84.17% |
| 39 | Xiaomi MIMO v2.6 Pro | 88.43% | 91.91% | 74.38% | 94.40% | 96.55% | 87.32% | 86.03% | 91.50% |
| 40 | Qwen 3.6 Max Preview | 88.42% | 83.61% | 67.96% | 91.85% | 98.07% | 99.15% | 89.87% | 93.72% |
| 41 | Aion 3.5 Mini (Reasoning, High) | 88.32% | 92.50% | 75.70% | 92.04% | 96.04% | 83.75% | 89.86% | 92.36% |
| 42 | Qwen 3.7 Flash (Reasoning) | 88.17% | 78.27% | 80.21% | 84.81% | 98.07% | 98.39% | 89.28% | 88.56% |
| 43 | Claude Opus 5 | 88.13% | 92.98% | 76.11% | 95.16% | 95.11% | 82.37% | 87.08% | 89.97% |
| 44 | GPT-5.4 Mini | 88.10% | 89.64% | 75.74% | 90.45% | 94.65% | 98.21% | 79.92% | 80.45% |
| 45 | GPT-5.4 Mini (Reasoning, Low) | 87.72% | 89.64% | 74.97% | 89.66% | 95.16% | 98.16% | 78.74% | 83.57% |
| 46 | Grok 4.5 (Reasoning, Low) | 87.45% | 88.61% | 73.15% | 91.72% | 95.00% | 89.93% | 86.31% | 90.94% |
| 47 | Grok 4.5 (Reasoning, High) | 87.43% | 88.48% | 74.99% | 92.01% | 93.96% | 89.17% | 85.97% | 94.12% |
| 48 | Kimi K3 (Reasoning, High) | 87.43% | 91.24% | 74.54% | 92.23% | 96.61% | 86.94% | 82.98% | 92.99% |
| 49 | Muse Glimmer 30B (Reasoning, Medium) | 87.42% | 90.38% | 61.62% | 94.53% | 96.53% | 96.49% | 84.94% | 88.81% |
| 50 | Muse Spark 1.2 (Reasoning, Medium) | 87.33% | 91.35% | 68.38% | 91.68% | 95.96% | 94.63% | 82.01% | 94.71% |
| 51 | Hy4 Preview (Reasoning, High) | 87.32% | 89.81% | 69.96% | 90.92% | 95.57% | 90.36% | 87.32% | 94.80% |
| 52 | Muse Spark 1.1 (Reasoning, Minimal) | 87.30% | 91.25% | 74.35% | 88.84% | 94.91% | 92.57% | 81.92% | 91.43% |
| 53 | GPT-5.1 | 87.20% | 88.21% | 73.76% | 87.44% | 93.57% | 93.87% | 86.39% | 90.73% |
| 54 | Qwen 3.5 397B A17B | 86.93% | 84.94% | 68.63% | 92.54% | 94.84% | 97.50% | 83.12% | 91.09% |
| 55 | GPT-5 | 86.87% | 92.05% | 62.58% | 91.30% | 92.06% | 93.65% | 89.59% | 91.48% |
| 56 | Kimi K3 (Reasoning, Low) | 86.51% | 89.73% | 75.28% | 90.50% | 95.04% | 82.70% | 85.82% | 92.07% |
| 57 | Grok 4.20 (Reasoning) | 86.25% | 84.92% | 74.25% | 83.26% | 94.23% | 92.86% | 87.99% | 90.87% |
| 58 | Aion 3.5 (Reasoning, High) | 86.16% | 91.42% | 74.14% | 89.96% | 96.01% | 80.27% | 85.17% | 91.90% |
| 59 | Z.AI GLM 5.3 Flash (Reasoning, Low) | 86.13% | 91.14% | 75.32% | 91.18% | 92.77% | 82.42% | 83.94% | 88.76% |
| 60 | Qwen 3.6 Flash | 86.02% | 82.10% | 67.13% | 86.74% | 96.80% | 97.83% | 85.50% | 89.31% |
| 61 | Qwen 3.6 35B | 85.97% | 81.28% | 65.03% | 89.07% | 96.65% | 97.44% | 86.36% | 87.66% |
| 62 | Claude Opus 4.8 (Reasoning, Low) | 85.86% | 90.38% | 74.07% | 87.92% | 94.06% | 85.67% | 83.05% | 91.89% |
| 63 | DeepSeek V4 Pro 0813 (Reasoning, High) | 85.72% | 89.68% | 68.88% | 90.25% | 95.04% | 90.05% | 80.44% | 90.21% |
| 64 | Kimi K2.6 | 85.47% | 86.79% | 68.21% | 92.60% | 95.55% | 91.73% | 77.97% | 92.57% |
| 65 | Gemini 3.1 Pro (Preview) | 85.44% | 80.73% | 63.42% | 87.20% | 95.06% | 97.27% | 88.95% | 94.08% |
| 66 | Qwen 3.7 Max | 85.39% | 82.38% | 61.91% | 88.93% | 95.46% | 96.29% | 87.36% | 94.55% |
| 67 | Gemini 4 Argon (Reasoning, Medium) | 85.27% | 86.48% | 79.32% | 79.65% | 94.01% | 86.33% | 85.82% | 93.81% |
| 68 | Claude Opus 4.8 (Reasoning) | 85.25% | 89.74% | 72.75% | 86.50% | 93.97% | 84.99% | 83.58% | 92.33% |
| 69 | Qwen 3.5 Plus (2026-04-20) | 85.18% | 82.23% | 64.62% | 85.08% | 97.91% | 96.76% | 84.50% | 89.79% |
| 70 | Grok 4.3 (Reasoning) | 85.11% | 85.99% | 58.99% | 90.45% | 94.05% | 93.58% | 87.57% | 90.99% |
| 71 | Qwen 3 235B A22B Instruct 2507 | 84.81% | 82.49% | 81.37% | 87.23% | 95.70% | 79.99% | 82.05% | 78.07% |
| 72 | Claude Opus 4.7 | 84.74% | 90.95% | 75.05% | 88.63% | 91.45% | 80.25% | 82.11% | 89.90% |
| 73 | Claude Opus 4.7 (Reasoning) | 84.73% | 91.22% | 72.71% | 88.14% | 92.86% | 79.84% | 83.59% | 92.53% |
| 74 | Z.AI GLM 5 Turbo | 84.66% | 86.97% | 75.38% | 93.23% | 87.80% | 86.19% | 78.37% | 93.29% |
| 75 | Aion 3.0 | 84.64% | 87.34% | 72.32% | 93.98% | 92.17% | 85.58% | 76.47% | 88.78% |
| 76 | Aion 3.0 Mini | 84.60% | 86.04% | 71.45% | 92.64% | 90.67% | 85.99% | 80.84% | 83.69% |
| 77 | MiniMax M3 | 84.57% | 89.40% | 70.33% | 93.86% | 90.97% | 84.36% | 78.49% | 90.45% |
| 78 | Claude Opus 4.6 (Reasoning) | 84.55% | 88.14% | 75.04% | 93.56% | 91.88% | 82.92% | 75.75% | 95.06% |
| 79 | Grok 4.3 | 84.51% | 84.86% | 62.78% | 87.13% | 95.49% | 91.18% | 85.60% | 78.00% |
| 80 | Qwen 3.5 9B | 84.35% | 87.56% | 53.95% | 91.95% | 96.66% | 96.17% | 79.82% | 84.05% |
| 81 | Claude Sonnet 4.5 | 84.19% | 84.67% | 78.05% | 89.63% | 89.39% | 86.97% | 76.41% | 87.54% |
| 82 | Xiaomi MIMO v2.6 Flash | 84.08% | 90.58% | 67.96% | 93.25% | 95.19% | 82.88% | 74.64% | 87.31% |
| 83 | Z.AI GLM 5.1 | 84.05% | 85.25% | 75.67% | 92.52% | 89.65% | 83.70% | 77.47% | 93.74% |
| 84 | DeepSeek V4 Flash 0731 (Reasoning, High) | 84.03% | 88.34% | 67.80% | 91.18% | 93.33% | 83.37% | 80.18% | 87.97% |
| 85 | Writer: Palmyra X5 | 83.95% | 82.97% | 83.01% | 84.42% | 94.98% | 79.92% | 78.38% | 78.11% |
| 86 | Z.AI GLM 5.2 (Reasoning, High) | 83.92% | 88.16% | 73.96% | 93.65% | 90.45% | 83.05% | 74.28% | 93.41% |
| 87 | Qwen 3.5 Flash | 83.81% | 84.96% | 58.83% | 87.13% | 94.71% | 95.18% | 82.05% | 85.66% |
| 88 | Claude Opus 4 | 83.79% | 82.47% | 77.58% | 89.08% | 90.42% | 84.86% | 78.30% | 87.22% |
| 89 | DeepSeek V4 Pro | 83.70% | 85.40% | 73.07% | 91.48% | 91.83% | 83.03% | 77.40% | 82.05% |
| 90 | Z.AI GLM 5 | 83.63% | 84.90% | 77.33% | 88.46% | 88.49% | 83.75% | 78.86% | 89.60% |
| 91 | Claude Opus 4.6 | 83.59% | 88.68% | 74.14% | 92.37% | 91.07% | 82.67% | 72.59% | 92.31% |
| 92 | Qwen 3.5 35B | 83.51% | 85.05% | 58.12% | 89.61% | 93.58% | 93.58% | 81.11% | 87.01% |
| 93 | Grok 4.20 | 83.44% | 83.35% | 74.09% | 77.13% | 90.27% | 89.30% | 86.50% | 81.21% |
| 94 | DeepSeek V4 Flash | 83.42% | 86.21% | 69.82% | 88.48% | 90.79% | 85.70% | 79.51% | 82.02% |
| 95 | Claude Sonnet 4.6 | 83.31% | 87.17% | 72.52% | 95.20% | 90.14% | 78.39% | 76.43% | 90.66% |
| 96 | Gemini 3.8 Flash (Reasoning, Medium) | 83.10% | 84.48% | 69.67% | 76.42% | 95.11% | 89.35% | 83.58% | 95.72% |
| 97 | Claude Sonnet 4.6 (Reasoning) | 83.09% | 88.70% | 71.90% | 94.92% | 89.19% | 78.68% | 75.13% | 93.64% |
| 98 | Gemini 3.7 Flash (Reasoning, Medium) | 83.06% | 82.47% | 69.21% | 78.74% | 95.63% | 89.30% | 82.98% | 94.85% |
| 99 | DeepSeek V4 Flash (Reasoning) | 83.03% | 86.00% | 70.46% | 88.10% | 92.27% | 81.27% | 80.09% | 88.06% |
| 100 | Qwen 3.5 122B | 83.02% | 84.79% | 56.46% | 86.76% | 94.27% | 95.52% | 80.31% | 90.32% |
| 101 | DeepSeek V4 Pro (Reasoning) | 82.99% | 86.09% | 73.38% | 84.25% | 91.70% | 84.41% | 78.10% | 89.28% |
| 102 | Qwen 3.6 27B | 82.81% | 80.96% | 63.42% | 78.25% | 96.75% | 96.84% | 80.66% | 88.33% |
| 103 | Gemini 3.5 Flash Lite (Reasoning) | 82.80% | 80.26% | 70.35% | 80.87% | 91.78% | 91.26% | 82.32% | 84.34% |
| 104 | Gemini 3.5 Flash Lite (Reasoning, Minimal) | 82.74% | 80.98% | 71.28% | 78.76% | 92.36% | 91.31% | 81.74% | 84.45% |
| 105 | o4 Mini High | 82.72% | 77.81% | 63.78% | 78.26% | 97.56% | 88.52% | 90.38% | 88.78% |
| 106 | Claude Sonnet 5 | 82.58% | 89.56% | 73.22% | 91.19% | 81.40% | 82.26% | 77.86% | 87.34% |
| 107 | Qwen 3.5 27B | 82.54% | 84.16% | 54.98% | 88.67% | 94.46% | 94.50% | 78.44% | 90.05% |
| 108 | ByteDance Seed 2.0 Lite | 82.35% | 93.39% | 58.28% | 85.96% | 90.78% | 89.48% | 76.22% | 84.27% |
| 109 | DeepSeek V3 (2025-03-24) | 82.34% | 81.15% | 73.86% | 83.31% | 95.33% | 79.89% | 80.49% | 79.93% |
| 110 | Mistral Medium 3.5 | 82.26% | 80.42% | 73.82% | 81.20% | 93.31% | 80.95% | 83.87% | 84.52% |
| 111 | Mistral Large 4.0 | 82.15% | 88.85% | 66.79% | 87.19% | 92.33% | 87.26% | 70.51% | 84.72% |
| 112 | o4 Mini | 82.04% | 78.02% | 63.95% | 78.87% | 97.30% | 87.50% | 86.60% | 86.56% |
| 113 | Mistral Large 2 | 81.86% | 81.31% | 72.75% | 83.10% | 91.94% | 80.68% | 81.36% | 81.50% |
| 114 | Claude Opus 4.5 | 81.71% | 84.65% | 76.11% | 88.91% | 83.27% | 83.45% | 73.89% | 89.60% |
| 115 | MiniMax M2.7 | 81.70% | 85.80% | 72.16% | 88.91% | 83.74% | 84.86% | 74.72% | 86.23% |
| 116 | Mistral Medium 3.1 | 81.70% | 82.67% | 74.30% | 81.51% | 94.59% | 77.16% | 79.96% | 76.08% |
| 117 | Mistral Small 4 (Reasoning) | 81.67% | 81.83% | 77.22% | 85.39% | 90.49% | 79.13% | 75.97% | 79.48% |
| 118 | Mistral Medium 3.5 (Reasoning) | 81.67% | 80.50% | 74.66% | 78.62% | 92.66% | 81.58% | 81.97% | 85.32% |
| 119 | ByteDance Seed 1.6 Flash | 81.51% | 90.73% | 70.46% | 79.96% | 95.11% | 77.28% | 75.52% | 70.92% |
| 120 | Claude Sonnet 5 (Reasoning, Low) | 81.43% | 89.01% | 74.99% | 87.10% | 82.23% | 81.53% | 73.75% | 90.16% |
| 121 | Kimi K2.5 | 81.35% | 84.62% | 70.84% | 86.21% | 93.06% | 86.00% | 67.36% | 90.86% |
| 122 | Claude Sonnet 5 (Reasoning) | 81.31% | 89.71% | 72.46% | 87.18% | 82.55% | 81.18% | 74.78% | 90.40% |
| 123 | Qwen 3 32B | 81.30% | 82.63% | 71.76% | 78.83% | 93.62% | 79.28% | 81.72% | 79.37% |
| 124 | GPT-4.1 | 81.24% | 82.39% | 69.77% | 76.96% | 93.57% | 80.93% | 83.80% | 86.82% |
| 125 | MiniMax M2.5 | 81.21% | 86.13% | 73.63% | 88.93% | 84.71% | 78.48% | 75.41% | 86.71% |
| 126 | Mistral Large 3 | 81.21% | 81.25% | 72.38% | 80.22% | 92.61% | 80.06% | 80.74% | 84.29% |
| 127 | Mistral Small 4 | 81.12% | 80.83% | 75.37% | 82.49% | 93.71% | 78.11% | 76.23% | 75.23% |
| 128 | Xiaomi MIMO v2.5 Pro | 81.08% | 83.10% | 68.44% | 79.81% | 92.48% | 86.42% | 76.24% | 86.05% |
| 129 | Gemini 2.5 Pro | 81.03% | 78.06% | 62.91% | 77.40% | 90.00% | 90.61% | 87.21% | 88.44% |
| 130 | Gemini 3.6 Flash (Reasoning) | 81.02% | 82.36% | 66.90% | 71.94% | 95.75% | 87.28% | 81.86% | 94.90% |
| 131 | GPT-5.4 Nano (Reasoning) | 80.97% | 86.40% | 69.53% | 85.05% | 84.09% | 88.12% | 72.62% | 80.02% |
| 132 | GPT-5.4 Nano (Reasoning, Low) | 80.93% | 86.94% | 71.57% | 83.10% | 84.72% | 88.80% | 70.47% | 77.46% |
| 133 | Hermes 3 405B | 80.92% | 78.30% | 69.94% | 68.75% | 93.61% | 92.33% | 82.60% | 80.80% |
| 134 | Mistral Large 4.0 (Reasoning) | 80.54% | 84.90% | 64.53% | 84.72% | 91.39% | 87.17% | 70.53% | 92.14% |
| 135 | GPT-5.4 Nano | 80.50% | 86.04% | 71.90% | 83.18% | 83.70% | 88.53% | 69.68% | 72.16% |
| 136 | GPT-5 Mini | 80.48% | 89.87% | 60.91% | 85.94% | 83.16% | 84.52% | 78.48% | 91.31% |
| 137 | GPT-5.2 | 80.36% | 88.60% | 67.99% | 80.71% | 87.55% | 90.94% | 66.36% | 89.45% |
| 138 | Thinking Machines Inkling | 80.27% | 85.74% | 69.67% | 85.63% | 92.99% | 86.32% | 61.24% | 73.96% |
| 139 | Aion 2.0 | 80.24% | 81.95% | 64.90% | 76.51% | 89.80% | 86.75% | 81.53% | 86.66% |
| 140 | Thinking Machines Inkling (Reasoning) | 80.19% | 84.62% | 71.97% | 86.71% | 89.46% | 85.97% | 62.40% | 90.78% |
| 141 | ByteDance Seed 2.0 Mini | 80.11% | 90.38% | 59.35% | 81.36% | 87.57% | 80.08% | 81.90% | 85.69% |
| 142 | DeepSeek V3.2 | 79.95% | 80.86% | 66.11% | 73.33% | 90.65% | 87.42% | 81.32% | 82.22% |
| 143 | Gemini 3.5 Flash (Reasoning) | 79.87% | 82.33% | 66.61% | 67.65% | 91.71% | 89.17% | 81.76% | 93.35% |
| 144 | Gemini 3.6 Flash (Reasoning, Minimal) | 79.25% | 82.36% | 66.51% | 64.68% | 94.74% | 86.78% | 80.44% | 86.68% |
| 145 | Claude Sonnet 4 | 79.21% | 79.20% | 76.88% | 82.84% | 78.57% | 84.00% | 73.80% | 87.64% |
| 146 | Xiaomi MIMO v2.5 | 79.16% | 82.45% | 66.42% | 77.40% | 88.79% | 82.42% | 77.49% | 83.95% |
| 147 | Ministral 3 14B | 79.11% | 80.51% | 74.94% | 74.59% | 92.74% | 73.18% | 78.70% | 70.45% |
| 148 | WizardLM 2 8x22b | 79.06% | 80.21% | 65.28% | 77.45% | 85.19% | 86.84% | 79.40% | 71.45% |
| 149 | Claude Haiku 4.5 | 78.96% | 84.13% | 75.25% | 82.25% | 81.79% | 78.41% | 71.93% | 83.36% |
| 150 | Z.AI GLM 4.7 | 78.89% | 83.61% | 61.49% | 71.01% | 91.38% | 84.23% | 81.62% | 87.67% |
| 151 | Z.AI GLM 4.6 | 78.86% | 79.06% | 63.85% | 74.32% | 84.37% | 87.76% | 83.84% | 87.64% |
| 152 | Gemma 3 27B | 78.79% | 75.45% | 71.72% | 64.44% | 88.37% | 88.98% | 83.82% | 75.70% |
| 153 | Gemini 3.5 Flash (Reasoning, Minimal) | 78.55% | 82.60% | 68.64% | 63.97% | 88.91% | 86.81% | 80.34% | 85.88% |
| 154 | ByteDance Seed 1.6 | 78.43% | 87.67% | 57.70% | 85.62% | 87.01% | 74.93% | 77.64% | 89.59% |
| 155 | Gemma 4 31B (Reasoning) | 78.13% | 78.69% | 60.53% | 80.03% | 85.05% | 84.94% | 79.53% | 89.64% |
| 156 | DeepSeek V3 (2024-12-26) | 77.88% | 76.53% | 66.31% | 69.79% | 91.42% | 83.43% | 79.81% | 82.62% |
| 157 | Cohere Command R+ (Aug. 2024) | 77.70% | 76.55% | 71.05% | 64.24% | 87.40% | 89.16% | 77.80% | 67.04% |
| 158 | Gemini 2.5 Flash | 77.57% | 73.23% | 65.37% | 70.64% | 83.37% | 94.77% | 78.05% | 80.61% |
| 159 | DeepSeek V3.1 | 77.45% | 78.57% | 62.92% | 68.61% | 86.19% | 84.49% | 83.91% | 82.35% |
| 160 | Hermes 3 70B | 77.41% | 77.97% | 73.70% | 59.25% | 87.04% | 88.06% | 78.42% | 69.74% |
| 161 | Z.AI GLM 4.7 Flash | 77.36% | 83.78% | 61.69% | 69.49% | 89.60% | 78.53% | 81.09% | 82.21% |
| 162 | Laguna XS 2.1 | 77.28% | 81.27% | 73.58% | 72.93% | 86.56% | 75.66% | 73.66% | 75.42% |
| 163 | Ministral 3 8B | 77.26% | 78.74% | 70.61% | 72.81% | 90.74% | 74.38% | 76.24% | 69.98% |
| 164 | DeepSeek-V2 Chat | 77.20% | 75.84% | 66.87% | 66.95% | 91.52% | 82.94% | 79.10% | 84.09% |
| 165 | Qwen 3.5 Plus (2026-02-15) | 77.07% | 81.88% | 67.63% | 61.06% | 87.42% | 92.52% | 71.92% | 86.17% |
| 166 | Ministral 8B | 76.87% | 78.83% | 69.62% | 76.29% | 89.29% | 73.33% | 73.81% | 63.77% |
| 167 | Mistral NeMO | 76.72% | 73.01% | 59.98% | 67.00% | 89.84% | 94.70% | 75.80% | 63.80% |
| 168 | Z.AI GLM 4.5 | 76.56% | 80.07% | 68.98% | 73.22% | 81.73% | 83.00% | 72.34% | 84.95% |
| 169 | Gemma 4 26B (Reasoning) | 76.38% | 76.28% | 59.82% | 81.41% | 78.14% | 82.87% | 79.74% | 89.02% |
| 170 | Gemini 3.1 Flash Lite (Reasoning) | 76.31% | 78.50% | 59.35% | 81.09% | 80.96% | 78.50% | 79.50% | 85.91% |
| 171 | Gemini 2.5 Flash (Reasoning) | 76.30% | 71.84% | 66.66% | 65.33% | 81.71% | 94.83% | 77.43% | 84.14% |
| 172 | Gemini 3.1 Flash Lite | 76.01% | 78.59% | 58.53% | 78.90% | 82.35% | 77.70% | 79.97% | 85.09% |
| 173 | Gemini 3 Flash (Preview, Reasoning) | 75.87% | 82.18% | 61.94% | 69.48% | 78.95% | 81.52% | 81.18% | 89.93% |
| 174 | Gemini 3.1 Flash Lite (Preview) | 75.78% | 78.70% | 59.92% | 75.91% | 82.09% | 78.41% | 79.66% | 85.41% |
| 175 | Gemma 4 31B | 75.59% | 79.15% | 62.09% | 70.38% | 83.35% | 81.81% | 76.74% | 85.23% |
| 176 | GPT-4o, Aug. 6th (temp=1) | 75.50% | 71.49% | 73.05% | 59.42% | 87.96% | 80.80% | 80.25% | 81.28% |
| 177 | Ministral 3B | 75.49% | 76.30% | 67.17% | 75.29% | 90.17% | 73.22% | 70.80% | 59.25% |
| 178 | Ministral 3 3B | 75.45% | 75.23% | 65.17% | 74.49% | 89.60% | 77.57% | 70.63% | 65.02% |
| 179 | Gemma 3 12B | 75.38% | 72.08% | 71.52% | 54.87% | 86.76% | 84.22% | 82.85% | 76.07% |
| 180 | Gemma 4 26B | 75.17% | 77.47% | 59.55% | 73.77% | 77.99% | 81.90% | 80.33% | 84.89% |
| 181 | Qwen 2.5 72B | 75.16% | 73.92% | 60.17% | 64.34% | 84.37% | 95.34% | 72.84% | 73.17% |
| 182 | Gemini 2.5 Flash Lite | 75.05% | 71.91% | 66.56% | 62.67% | 79.16% | 93.78% | 76.25% | 79.91% |
| 183 | Gemini 3 Flash (Preview) | 75.04% | 83.34% | 61.35% | 64.78% | 84.15% | 76.32% | 80.32% | 85.47% |
| 184 | Thinking Machines Inkling Small (Reasoning, High) | 74.85% | 86.96% | 63.95% | 82.72% | 81.02% | 80.59% | 53.85% | 87.39% |
| 185 | Z.AI GLM 4.5 Air | 74.61% | 80.00% | 67.71% | 65.43% | 81.79% | 84.17% | 68.55% | 80.74% |
| 186 | GPT-4.1 Mini | 74.52% | 72.34% | 68.20% | 57.72% | 87.91% | 80.59% | 80.38% | 81.40% |
| 187 | GPT-4o Mini (temp=1) | 74.37% | 70.73% | 71.50% | 53.50% | 91.21% | 81.11% | 78.20% | 77.82% |
| 188 | Cydonia 24B V4.1 | 74.19% | 77.25% | 78.35% | 62.72% | 77.12% | 77.93% | 71.78% | 72.68% |
| 189 | Laguna S 2.1 | 74.16% | 79.33% | 73.87% | 68.01% | 81.48% | 75.38% | 66.91% | 79.99% |
| 190 | Arcee AI: Trinity Mini | 74.01% | 75.18% | 56.69% | 63.42% | 85.69% | 83.32% | 79.76% | 67.68% |
| 191 | GPT-4o, Aug. 6th (temp=0) | 73.65% | 72.11% | 58.95% | 63.20% | 84.90% | 93.39% | 69.33% | 82.18% |
| 192 | GPT-4o Mini (temp=0) | 73.10% | 70.53% | 59.43% | 56.91% | 88.83% | 92.33% | 70.56% | 76.86% |
| 193 | Llama 3.1 70B | 72.78% | 74.82% | 65.53% | 70.76% | 71.74% | 82.99% | 70.83% | 77.41% |
| 194 | Gemma 3 4B | 72.10% | 69.45% | 68.92% | 52.95% | 83.05% | 78.08% | 80.17% | 66.33% |
| 195 | Mistral Small 3.2 24B | 71.87% | 72.69% | 47.89% | 68.93% | 84.70% | 88.45% | 68.55% | 77.36% |
| 196 | GPT-4.1 Nano | 71.81% | 68.54% | 67.03% | 55.61% | 82.65% | 76.30% | 80.71% | 69.90% |
| 197 | Gemini 2.5 Flash Lite (Reasoning) | 71.64% | 71.77% | 65.69% | 65.73% | 72.59% | 87.33% | 66.75% | 83.10% |
| 198 | Nemotron 3 Super | 69.75% | 76.58% | 57.81% | 67.38% | 70.69% | 81.62% | 64.41% | 81.69% |
| 199 | Inception Mercury 2 | 68.31% | 74.37% | 50.37% | 59.94% | 67.97% | 89.69% | 67.53% | 81.99% |
| 200 | GPT-OSS 120B | 67.85% | 75.69% | 50.96% | 55.74% | 70.06% | 84.17% | 70.48% | 84.81% |
| 201 | GPT-5 Nano | 67.04% | 84.30% | 60.85% | 62.32% | 66.77% | 77.89% | 50.07% | 80.16% |
| 202 | Nemotron 3 Nano | 65.87% | 74.00% | 52.17% | 60.70% | 59.84% | 84.19% | 64.30% | 74.50% |