Model Leaderboard

All models ranked by their AI-isms subcategory score.

# Model AI-isms Creative Writing Overall
1 ByteDance Seed 2.0 Lite 93.39% 82.35% 84.80%
2 GPT-5.5 (Reasoning) 92.18% 90.26% 92.98%
3 GPT-5 92.05% 86.87% 91.93%
4 GPT-5.4 (Reasoning) 91.56% 91.17% 93.24%
5 GPT-5.5 91.50% 90.39% 89.09%
6 GPT-5.5 (Reasoning, Low) 91.49% 90.24% 92.59%
7 Claude Opus 4.7 (Reasoning) 91.22% 84.73% 93.23%
8 GPT-5.4 91.06% 90.94% 84.32%
9 GPT-5.4 (Reasoning, Low) 91.02% 90.51% 91.41%
10 Claude Opus 4.7 90.95% 84.74% 89.93%
11 GPT-5.4 Mini (Reasoning) 90.82% 88.66% 90.65%
12 ByteDance Seed 1.6 Flash 90.73% 81.51% 73.27%
13 ByteDance Seed 2.0 Mini 90.38% 80.11% 86.91%
14 Claude Opus 4.8 (Reasoning, Low) 90.38% 85.86% 92.14%
15 GPT-5 Mini 89.87% 80.48% 92.62%
16 Claude Opus 4.8 (Reasoning) 89.74% 85.25% 92.22%
17 GPT-5.4 Mini 89.64% 88.10% 82.43%
18 GPT-5.4 Mini (Reasoning, Low) 89.64% 87.72% 85.75%
19 MiniMax M3 89.40% 84.57% 90.88%
20 Claude Sonnet 4.6 (Reasoning) 88.70% 83.09% 93.66%
21 Claude Opus 4.6 88.68% 83.59% 92.35%
22 GPT-5.2 88.60% 80.36% 90.26%
23 GPT-5.1 88.21% 87.20% 92.54%
24 Claude Opus 4.6 (Reasoning) 88.14% 84.55% 95.02%
25 ByteDance Seed 1.6 87.67% 78.43% 90.70%
26 Qwen 3.5 9B 87.56% 84.35% 86.05%
27 Claude Sonnet 4.6 87.17% 83.31% 91.15%
28 Z.AI GLM 5 Turbo 86.97% 84.66% 94.27%
29 GPT-5.4 Nano (Reasoning, Low) 86.94% 80.93% 79.48%
30 MoonshotAI: Kimi K2.6 86.79% 85.47% 92.31%
31 GPT-5.4 Nano (Reasoning) 86.40% 80.97% 81.36%
32 DeepSeek V4 Flash 86.21% 83.42% 82.02%
33 MiniMax M2.5 86.13% 81.21% 88.71%
34 DeepSeek V4 Pro (Reasoning) 86.09% 82.99% 90.10%
35 GPT-5.4 Nano 86.04% 80.50% 74.40%
36 DeepSeek V4 Flash (Reasoning) 86.00% 83.03% 89.01%
37 Grok 4.3 (Reasoning) 85.99% 85.11% 93.60%
38 MiniMax M2.7 85.80% 81.70% 89.10%
39 DeepSeek V4 Pro 85.40% 83.70% 82.63%
40 Grok 4.20 (Beta, Reasoning) 85.37% 84.50% 91.49%
41 Z.AI GLM 5.1 85.25% 84.05% 94.37%
42 Qwen 3.5 35B 85.05% 83.51% 88.00%
43 Qwen 3.5 Flash 84.96% 83.81% 86.38%
44 Qwen 3.5 397B A17B 84.94% 86.93% 91.73%
45 Grok 4.20 (Reasoning) 84.92% 86.25% 91.39%
46 Z.AI GLM 5 84.90% 83.63% 91.23%
47 Grok 4.3 84.86% 84.51% 78.66%
48 Qwen 3.5 122B 84.79% 83.02% 91.53%
49 Claude Sonnet 4.5 84.67% 84.19% 88.03%
50 Claude Opus 4.5 84.65% 81.71% 89.69%
51 MoonshotAI: Kimi K2.5 84.62% 81.35% 91.04%
52 GPT-5 Nano 84.30% 67.04% 82.60%
53 Qwen 3.5 27B 84.16% 82.54% 90.85%
54 Claude Haiku 4.5 84.13% 78.96% 85.14%
55 Z.AI GLM 4.7 Flash 83.78% 77.36% 84.82%
56 Qwen3.6 Max Preview 83.61% 88.42% 94.54%
57 Z.AI GLM 4.7 83.61% 78.89% 88.69%
58 Grok 4.20 83.35% 83.44% 81.70%
59 Gemini 3 Flash (Preview) 83.34% 75.04% 85.35%
60 Gemini 3 Pro (Preview) 83.34% 77.77% 88.79%
61 Xiaomi MIMO v2.5 Pro 83.10% 81.08% 87.36%
62 Writer: Palmyra X5 82.97% 83.95% 79.57%
63 Mistral Medium 3.1 82.67% 81.70% 77.83%
64 Grok 4.20 (Beta) 82.65% 82.80% 83.85%
65 Qwen 3 32B 82.63% 81.30% 82.21%
66 Gemini 3.5 Flash (Reasoning, Minimal) 82.60% 78.55% 86.47%
67 Qwen3 235B A22B Instruct 2507 82.49% 84.81% 80.10%
68 Claude Opus 4 82.47% 83.79% 87.69%
69 Xiaomi MIMO v2.5 82.45% 79.16% 85.05%
70 GPT-4.1 82.39% 81.24% 88.68%
71 Qwen3.7 Max 82.38% 85.39% 95.75%
72 Gemini 3.5 Flash (Reasoning) 82.33% 79.87% 94.08%
73 Qwen 3.5 Plus (2026-04-20) 82.23% 85.18% 91.51%
74 Gemini 3 Flash (Preview, Reasoning) 82.18% 75.87% 90.50%
75 Qwen 3.6 Flash 82.10% 86.02% 90.65%
76 Skyfall 36B V2 82.10% 83.32% 65.76%
77 Rocinante 12B 81.99% 81.94% 54.54%
78 Aion 2.0 81.95% 80.24% 89.21%
79 Qwen 3.5 Plus (2026-02-15) 81.88% 77.07% 85.96%
80 Mistral Small 4 (Reasoning) 81.83% 81.67% 82.39%
81 Mistral Large 81.60% 82.02% 80.15%
82 Stealth: Hunter Alpha 81.32% 79.18% 87.34%
83 Mistral Large 2 81.31% 81.86% 82.41%
84 Qwen 3.6 35B 81.28% 85.97% 89.05%
85 Mistral Large 3 81.25% 81.21% 85.43%
86 DeepSeek V3 (2025-03-24) 81.15% 82.34% 81.99%
87 Stealth: Healer Alpha 81.02% 78.28% 85.93%
88 Qwen 3.6 27B 80.96% 82.81% 89.72%
89 Mistral Small Creative 80.94% 80.29% 73.27%
90 DeepSeek V3.2 80.86% 79.95% 82.25%
91 Mistral Small 4 80.83% 81.12% 76.46%
92 Gemini 3.1 Pro (Preview) 80.73% 85.44% 94.37%
93 Claude 3.7 Sonnet 80.64% 76.31% 83.39%
94 Ministral 3 14B 80.51% 79.11% 72.54%
95 WizardLM 2 8x22b 80.21% 79.06% 71.06%
96 Z.AI GLM 4.5 80.07% 76.56% 86.27%
97 Z.AI GLM 4.5 Air 80.00% 74.61% 83.12%
98 Claude Sonnet 4 79.20% 79.21% 88.72%
99 Gemma 4 31B 79.15% 75.59% 86.91%
100 Z.AI GLM 4.6 79.06% 78.86% 89.11%
101 Ministral 8B 78.83% 76.87% 64.87%
102 Claude 3.5 Sonnet 78.78% 78.69% 84.24%
103 Ministral 3 8B 78.74% 77.26% 71.76%
104 Gemini 3.1 Flash Lite (Preview) 78.70% 75.78% 85.87%
105 Gemma 4 31B (Reasoning) 78.69% 78.13% 91.71%
106 Arcee AI: Trinity Large (Preview) 78.62% 75.26% 73.33%
107 Gemini 3.1 Flash Lite 78.59% 76.01% 85.75%
108 DeepSeek V3.1 78.57% 77.45% 82.39%
109 Gemini 3.1 Flash Lite (Reasoning) 78.50% 76.31% 86.41%
110 Llama 3.1 8B 78.41% 76.54% 63.35%
111 Hermes 3 405B 78.30% 80.92% 82.86%
112 Gemini 2.5 Pro 78.06% 81.03% 88.53%
113 Grok 4.1 Fast 78.06% 82.14% 89.55%
114 Grok 4 78.04% 77.34% 88.12%
115 o4 Mini 78.02% 82.04% 88.35%
116 Hermes 3 70B 77.97% 77.41% 72.57%
117 o4 Mini High 77.81% 82.72% 90.29%
118 Gemma 4 26B 77.47% 75.17% 85.84%
119 Cydonia 24B V4.1 77.25% 74.19% 75.09%
120 Grok 4 Fast 76.89% 77.03% 86.15%
121 Nemotron 3 Super 76.58% 69.75% 84.56%
122 Cohere Command R+ (Aug. 2024) 76.55% 77.70% 69.03%
123 DeepSeek V3 (2024-12-26) 76.53% 77.88% 83.68%
124 LFM2 24B 76.37% 78.10% 58.77%
125 Ministral 3B 76.30% 75.49% 61.29%
126 Gemma 4 26B (Reasoning) 76.28% 76.38% 91.49%
127 DeepSeek-V2 Chat 75.84% 77.20% 84.83%
128 GPT-OSS 120B 75.69% 67.85% 86.44%
129 Inception Mercury 75.55% 69.99% 79.50%
130 Gemma 3 27B 75.45% 78.79% 77.85%
131 Ministral 3 3B 75.23% 75.45% 67.22%
132 Arcee AI: Trinity Mini 75.18% 74.01% 70.90%
133 Llama 3.1 70B 74.82% 72.78% 78.40%
134 Inception Mercury 2 74.37% 68.31% 83.85%
135 Stealth: Aurora Alpha 74.28% 67.54% 83.79%
136 Nemotron 3 Nano 74.00% 65.87% 77.73%
137 Qwen 2.5 72B 73.92% 75.16% 75.46%
138 Gemini 2.5 Flash 73.23% 77.57% 80.60%
139 Mistral NeMO 73.01% 76.72% 65.04%
140 GPT-4o, May 13th (temp=0) 72.87% 74.89% 85.36%
141 Mistral Small 3.2 24B 72.69% 71.87% 78.58%
142 GPT-4.1 Mini 72.34% 74.52% 83.20%
143 GPT-4o, May 13th (temp=1) 72.12% 75.88% 83.80%
144 GPT-4o, Aug. 6th (temp=0) 72.11% 73.65% 82.45%
145 Gemma 3 12B 72.08% 75.38% 78.41%
146 Gemini 2.5 Flash Lite 71.91% 75.05% 81.08%
147 Gemini 2.5 Flash (Reasoning) 71.84% 76.30% 86.51%
148 Llama 3.1 Nemotron 70B 71.82% 71.71% 74.70%
149 Gemini 2.5 Flash Lite (Reasoning) 71.77% 71.64% 85.75%
150 GPT-4o, Aug. 6th (temp=1) 71.49% 75.50% 82.62%
151 Claude 3 Haiku 70.91% 74.53% 71.19%
152 GPT-4o Mini (temp=1) 70.73% 74.37% 79.08%
153 GPT-4o Mini (temp=0) 70.53% 73.10% 78.29%
154 Gemma 3 4B 69.45% 72.10% 68.57%
155 GPT-4.1 Nano 68.54% 71.81% 71.94%