Mechanical Style

Subcategory of Creative Writing. 146 models scored.

Model Leaderboard

All models ranked by their Mechanical Style subcategory score.

# Model Mechanical Style Creative Writing Overall
1 GPT-5.6 Sol (Reasoning) 99.81% 91.03% 95.15%
2 GPT-5.6 Sol 99.63% 91.04% 90.76%
3 GPT-5.6 Luna (Reasoning) 99.54% 90.10% 90.61%
4 GPT-5.6 Terra 99.54% 90.92% 88.23%
5 GPT-5.6 Luna 99.39% 89.69% 85.06%
6 GPT-5.5 (Reasoning) 99.27% 90.26% 93.72%
7 GPT-5.5 (Reasoning, Low) 99.26% 90.24% 92.51%
8 GPT-5.5 99.17% 90.39% 89.37%
9 Qwen3.6 Max Preview 99.15% 88.42% 93.72%
10 GPT-5.6 Terra (Reasoning) 99.04% 90.72% 92.49%
11 GPT-5.4 Mini (Reasoning) 98.70% 88.66% 89.82%
12 GPT-5.4 Mini 98.21% 88.10% 80.45%
13 GPT-5.4 Mini (Reasoning, Low) 98.16% 87.72% 83.57%
14 GPT-5.4 (Reasoning) 97.97% 91.17% 93.85%
15 Qwen 3.6 Flash 97.83% 86.02% 89.31%
16 GPT-5.4 (Reasoning, Low) 97.75% 90.51% 90.91%
17 Qwen 3.5 397B A17B 97.50% 86.93% 91.09%
18 GPT-5.4 97.49% 90.94% 84.31%
19 Qwen 3.6 35B 97.44% 85.97% 87.66%
20 Gemini 3.1 Pro (Preview) 97.27% 85.44% 94.08%
21 Qwen 3.6 27B 96.84% 82.81% 88.33%
22 Qwen 3.5 Plus (2026-04-20) 96.76% 85.18% 89.79%
23 Qwen3.7 Max 96.29% 85.39% 94.55%
24 Qwen 3.5 9B 96.17% 84.35% 84.05%
25 Qwen 3.5 122B 95.52% 83.02% 90.32%
26 Qwen 2.5 72B 95.34% 75.16% 73.17%
27 Qwen 3.5 Flash 95.18% 83.81% 85.66%
28 Gemini 2.5 Flash (Reasoning) 94.83% 76.30% 84.14%
29 Gemini 2.5 Flash 94.77% 77.57% 80.61%
30 Mistral NeMO 94.70% 76.72% 63.80%
31 Qwen 3.5 27B 94.50% 82.54% 90.05%
32 GPT-5.1 93.87% 87.20% 90.73%
33 Gemini 2.5 Flash Lite 93.78% 75.05% 79.91%
34 GPT-5 93.65% 86.87% 91.48%
35 Grok 4.3 (Reasoning) 93.58% 85.11% 90.99%
36 Qwen 3.5 35B 93.58% 83.51% 87.01%
37 GPT-4o, Aug. 6th (temp=0) 93.39% 73.65% 82.18%
38 Grok 4.20 (Reasoning) 92.86% 86.25% 90.87%
39 Qwen 3.5 Plus (2026-02-15) 92.52% 77.07% 86.17%
40 GPT-4o Mini (temp=0) 92.33% 73.10% 76.86%
41 Hermes 3 405B 92.33% 80.92% 80.80%
42 MoonshotAI: Kimi K2.6 91.73% 85.47% 92.57%
43 Grok 4.3 91.18% 84.51% 78.00%
44 GPT-5.2 90.94% 80.36% 89.45%
45 Gemini 2.5 Pro 90.61% 81.03% 88.44%
46 Grok 4.5 (Reasoning, Low) 89.93% 87.45% 90.94%
47 Inception Mercury 2 89.69% 68.31% 81.99%
48 ByteDance Seed 2.0 Lite 89.48% 82.35% 84.27%
49 Grok 4.20 89.30% 83.44% 81.21%
50 Gemini 3.5 Flash (Reasoning) 89.17% 79.87% 93.35%
51 Grok 4.5 (Reasoning, High) 89.17% 87.43% 94.12%
52 Cohere Command R+ (Aug. 2024) 89.16% 77.70% 67.04%
53 Gemma 3 27B 88.98% 78.79% 75.70%
54 GPT-5.4 Nano (Reasoning, Low) 88.80% 80.93% 77.46%
55 GPT-5.4 Nano 88.53% 80.50% 72.16%
56 o4 Mini High 88.52% 82.72% 88.78%
57 Mistral Small 3.2 24B 88.45% 71.87% 77.36%
58 GPT-5.4 Nano (Reasoning) 88.12% 80.97% 80.02%
59 Hermes 3 70B 88.06% 77.41% 69.74%
60 Z.AI GLM 4.6 87.76% 78.86% 87.64%
61 o4 Mini 87.50% 82.04% 86.56%
62 DeepSeek V3.2 87.42% 79.95% 82.22%
63 Gemini 2.5 Flash Lite (Reasoning) 87.33% 71.64% 83.10%
64 Claude Sonnet 4.5 86.97% 84.19% 87.54%
65 WizardLM 2 8x22b 86.84% 79.06% 71.45%
66 Gemini 3.5 Flash (Reasoning, Minimal) 86.81% 78.55% 85.88%
67 Aion 2.0 86.75% 80.24% 86.66%
68 Xiaomi MIMO v2.5 Pro 86.42% 81.08% 86.05%
69 Z.AI GLM 5 Turbo 86.19% 84.66% 93.29%
70 MoonshotAI: Kimi K2.5 86.00% 81.35% 90.86%
71 Aion 3.0 Mini 85.99% 84.60% 83.69%
72 DeepSeek V4 Flash 85.70% 83.42% 82.02%
73 Claude Opus 4.8 (Reasoning, Low) 85.67% 85.86% 91.89%
74 Aion 3.0 85.58% 84.64% 88.78%
75 Claude Opus 4.8 (Reasoning) 84.99% 85.25% 92.33%
76 Gemma 4 31B (Reasoning) 84.94% 78.13% 89.64%
77 MiniMax M2.7 84.86% 81.70% 86.23%
78 Claude Opus 4 84.86% 83.79% 87.22%
79 GPT-5 Mini 84.52% 80.48% 91.31%
80 DeepSeek V3.1 84.49% 77.45% 82.35%
81 DeepSeek V4 Pro (Reasoning) 84.41% 82.99% 89.28%
82 MiniMax M3 84.36% 84.57% 90.45%
83 Z.AI GLM 4.7 84.23% 78.89% 87.67%
84 Gemma 3 12B 84.22% 75.38% 76.07%
85 Nemotron 3 Nano 84.19% 65.87% 74.50%
86 GPT-OSS 120B 84.17% 67.85% 84.81%
87 Z.AI GLM 4.5 Air 84.17% 74.61% 80.74%
88 Claude Sonnet 4 84.00% 79.21% 87.64%
89 Z.AI GLM 5 83.75% 83.63% 89.60%
90 Z.AI GLM 5.1 83.70% 84.05% 93.74%
91 Claude Opus 4.5 83.45% 81.71% 89.60%
92 DeepSeek V3 (2024-12-26) 83.43% 77.88% 82.62%
93 Arcee AI: Trinity Mini 83.32% 74.01% 67.68%
94 Z.AI GLM 5.2 (Reasoning, High) 83.05% 83.92% 93.41%
95 DeepSeek V4 Pro 83.03% 83.70% 82.05%
96 Z.AI GLM 4.5 83.00% 76.56% 84.95%
97 Llama 3.1 70B 82.99% 72.78% 77.41%
98 DeepSeek-V2 Chat 82.94% 77.20% 84.09%
99 Claude Opus 4.6 (Reasoning) 82.92% 84.55% 95.06%
100 Gemma 4 26B (Reasoning) 82.87% 76.38% 89.02%
101 Claude Opus 4.6 82.67% 83.59% 92.31%
102 Xiaomi MIMO v2.5 82.42% 79.16% 83.95%
103 Claude Sonnet 5 82.26% 82.58% 87.34%
104 Gemma 4 26B 81.90% 75.17% 84.89%
105 Gemma 4 31B 81.81% 75.59% 85.23%
106 Nemotron 3 Super 81.62% 69.75% 81.69%
107 Claude Sonnet 5 (Reasoning, Low) 81.53% 81.43% 90.16%
108 Gemini 3 Flash (Preview, Reasoning) 81.52% 75.87% 89.93%
109 DeepSeek V4 Flash (Reasoning) 81.27% 83.03% 88.06%
110 Claude Sonnet 5 (Reasoning) 81.18% 81.31% 90.40%
111 GPT-4o Mini (temp=1) 81.11% 74.37% 77.82%
112 GPT-4.1 80.93% 81.24% 86.82%
113 GPT-4o, Aug. 6th (temp=1) 80.80% 75.50% 81.28%
114 Mistral Large 2 80.68% 81.86% 81.50%
115 GPT-4.1 Mini 80.59% 74.52% 81.40%
116 Claude Opus 4.7 80.25% 84.74% 89.90%
117 ByteDance Seed 2.0 Mini 80.08% 80.11% 85.69%
118 Mistral Large 3 80.06% 81.21% 84.29%
119 Qwen3 235B A22B Instruct 2507 79.99% 84.81% 78.07%
120 Writer: Palmyra X5 79.92% 83.95% 78.11%
121 DeepSeek V3 (2025-03-24) 79.89% 82.34% 79.93%
122 Claude Opus 4.7 (Reasoning) 79.84% 84.73% 92.53%
123 Qwen 3 32B 79.28% 81.30% 79.37%
124 Mistral Small 4 (Reasoning) 79.13% 81.67% 79.48%
125 Claude Sonnet 4.6 (Reasoning) 78.68% 83.09% 93.64%
126 Z.AI GLM 4.7 Flash 78.53% 77.36% 82.21%
127 Gemini 3.1 Flash Lite (Reasoning) 78.50% 76.31% 85.91%
128 MiniMax M2.5 78.48% 81.21% 86.71%
129 Gemini 3.1 Flash Lite (Preview) 78.41% 75.78% 85.41%
130 Claude Haiku 4.5 78.41% 78.96% 83.36%
131 Claude Sonnet 4.6 78.39% 83.31% 90.66%
132 Mistral Small 4 78.11% 81.12% 75.23%
133 Gemma 3 4B 78.08% 72.10% 66.33%
134 Cydonia 24B V4.1 77.93% 74.19% 72.68%
135 GPT-5 Nano 77.89% 67.04% 80.16%
136 Gemini 3.1 Flash Lite 77.70% 76.01% 85.09%
137 Ministral 3 3B 77.57% 75.45% 65.02%
138 ByteDance Seed 1.6 Flash 77.28% 81.51% 70.92%
139 Mistral Medium 3.1 77.16% 81.70% 76.08%
140 Gemini 3 Flash (Preview) 76.32% 75.04% 85.47%
141 GPT-4.1 Nano 76.30% 71.81% 69.90%
142 ByteDance Seed 1.6 74.93% 78.43% 89.59%
143 Ministral 3 8B 74.38% 77.26% 69.98%
144 Ministral 8B 73.33% 76.87% 63.77%
145 Ministral 3B 73.22% 75.49% 59.25%
146 Ministral 3 14B 73.18% 79.11% 70.45%