Word Counting

Subcategory of Utility. 118 models scored.

Model Leaderboard

All models ranked by their Word Counting subcategory score.

# Model Word Counting Utility Overall
1 Gemini 3.1 Pro (Preview) 99.56% 99.91% 94.37%
2 GPT-5 Mini 97.93% 98.39% 92.62%
3 GPT-5 96.13% 93.53% 91.93%
4 o4 Mini High 95.97% 98.67% 90.29%
5 Z.AI GLM 5 Turbo 95.82% 96.36% 94.27%
6 Claude Opus 4.6 (Reasoning) 94.65% 98.93% 95.02%
7 MiniMax M2.7 94.05% 95.50% 89.10%
8 Inception Mercury 2 93.76% 92.86% 83.85%
9 Nemotron 3 Super 93.51% 95.29% 84.56%
10 o4 Mini 90.72% 96.31% 88.35%
11 Claude Sonnet 4.6 (Reasoning) 90.42% 97.88% 93.66%
12 Gemini 3.1 Flash Lite (Preview) 89.38% 94.00% 85.87%
13 Stealth: Aurora Alpha 88.58% 92.59% 83.79%
14 Qwen 3.5 397B A17B 87.52% 97.50% 91.73%
15 Nemotron 3 Nano 86.70% 86.00% 77.73%
16 GPT-5 Nano 86.54% 93.91% 82.60%
17 MiniMax M2.5 86.03% 90.42% 88.71%
18 Gemini 3 Flash (Preview, Reasoning) 86.00% 97.20% 90.50%
19 Qwen 3.5 27B 85.82% 95.67% 90.85%
20 GPT-5.1 85.32% 95.33% 92.54%
21 Inception Mercury 85.14% 87.38% 79.50%
22 GPT-5.4 Nano (Reasoning) 84.86% 93.34% 81.36%
23 GPT-5.4 (Reasoning) 84.46% 96.89% 93.24%
24 GPT-5.2 84.32% 96.22% 90.26%
25 Gemini 3 Flash (Preview) 83.97% 86.39% 85.35%
26 Claude Opus 4.6 83.65% 90.72% 92.35%
27 MoonshotAI: Kimi K2.5 83.17% 96.63% 91.04%
28 GPT-4o Mini (temp=0) 82.65% 81.43% 78.29%
29 Qwen 3.5 35B 82.08% 96.42% 88.00%
30 Qwen 3.5 122B 81.81% 96.36% 91.53%
31 GPT-5.4 Mini (Reasoning) 81.72% 94.44% 90.65%
32 GPT-4.1 81.33% 90.57% 88.68%
33 Claude Opus 4.5 80.83% 89.84% 89.69%
34 Gemini 3 Pro (Preview) 80.75% 96.14% 88.79%
35 Qwen 3.5 Flash 80.57% 96.11% 86.38%
36 Z.AI GLM 4.7 78.57% 94.31% 88.69%
37 GPT-4o Mini (temp=1) 78.18% 82.16% 79.08%
38 GPT-4o, Aug. 6th (temp=1) 77.45% 82.44% 82.62%
39 Grok 4.20 (Beta, Reasoning) 77.05% 95.41% 91.49%
40 GPT-4o, Aug. 6th (temp=0) 77.04% 82.11% 82.45%
41 GPT-5.4 (Reasoning, Low) 76.62% 95.32% 91.41%
42 Z.AI GLM 5 75.55% 94.11% 91.23%
43 Grok 4 75.34% 89.67% 88.12%
44 GPT-5.4 Mini (Reasoning, Low) 75.01% 88.49% 85.75%
45 Qwen 3.5 9B 74.59% 94.02% 86.05%
46 Claude Opus 4 74.37% 88.81% 87.69%
47 ByteDance Seed 1.6 74.14% 90.83% 90.70%
48 ByteDance Seed 2.0 Mini 73.92% 91.88% 86.91%
49 ByteDance Seed 2.0 Lite 73.68% 92.23% 84.80%
50 GPT-5.4 Nano (Reasoning, Low) 73.42% 91.42% 79.48%
51 Claude Sonnet 4.6 71.94% 88.52% 91.15%
52 Z.AI GLM 4.7 Flash 68.93% 88.98% 84.82%
53 GPT-4.1 Mini 66.59% 82.30% 83.20%
54 Gemini 2.5 Pro 66.47% 92.18% 88.53%
55 GPT-5.4 Mini 66.13% 79.37% 82.43%
56 Claude 3.7 Sonnet 66.00% 62.54% 83.39%
57 GPT-4o, May 13th (temp=1) 65.90% 80.69% 83.80%
58 GPT-4o, May 13th (temp=0) 65.85% 83.13% 85.36%
59 Aion 2.0 65.48% 90.91% 89.21%
60 GPT-4.1 Nano 64.57% 68.45% 71.94%
61 Mistral Medium 3.1 64.40% 80.13% 77.83%
62 GPT-5.4 64.23% 81.95% 84.32%
63 Stealth: Healer Alpha 61.71% 82.30% 85.93%
64 Grok 4 Fast 61.39% 76.76% 86.15%
65 Grok 4.1 Fast 60.51% 84.12% 89.55%
66 Grok 4.20 (Beta) 59.16% 82.15% 83.85%
67 Claude Sonnet 4.5 58.93% 83.78% 88.03%
68 DeepSeek V3 (2025-03-24) 58.32% 80.62% 81.99%
69 DeepSeek V3 (2024-12-26) 58.30% 81.87% 83.68%
70 Llama 3.1 8B 56.94% 74.82% 63.37%
71 Llama 3.1 70B 56.62% 81.03% 78.40%
72 Mistral Small 4 (Reasoning) 56.00% 85.61% 82.39%
73 Stealth: Hunter Alpha 55.55% 84.63% 87.34%
74 Gemini 2.5 Flash Lite (Reasoning) 55.13% 89.63% 85.75%
75 Claude Sonnet 4 53.92% 84.02% 88.72%
76 Claude 3.5 Sonnet 53.50% 76.75% 84.24%
77 Gemini 2.5 Flash Lite 53.37% 80.14% 81.08%
78 Mistral Large 3 52.60% 84.91% 85.43%
79 Claude Haiku 4.5 51.60% 72.48% 85.14%
80 Qwen 2.5 72B 51.43% 76.43% 75.46%
81 Hermes 3 405B 51.42% 69.02% 82.86%
82 ByteDance Seed 1.6 Flash 48.55% 84.16% 73.27%
83 GPT-5.4 Nano 47.90% 78.57% 74.40%
84 Ministral 3 14B 47.57% 79.03% 72.54%
85 Z.AI GLM 4.6 47.21% 88.58% 89.11%
86 DeepSeek V3.1 47.10% 76.65% 82.39%
87 Llama 3.1 Nemotron 70B 47.07% 88.31% 74.70%
88 Gemma 3 12B 46.84% 79.28% 78.41%
89 Mistral Small 4 46.07% 78.28% 76.46%
90 Claude 3.5 Haiku 46.03% 82.57% 83.73%
91 Gemini 2.5 Flash (Reasoning) 46.02% 82.25% 86.51%
92 DeepSeek V3.2 45.81% 81.58% 82.25%
93 Gemma 3 27B 45.66% 76.82% 77.85%
94 LFM2 24B 44.98% 69.48% 58.77%
95 Qwen 3.5 Plus (2026-02-15) 44.24% 86.65% 85.96%
96 Arcee AI: Trinity Large (Preview) 43.69% 60.74% 73.33%
97 Mistral Small 3.2 24B 42.38% 73.17% 78.60%
98 DeepSeek-V2 Chat 41.72% 83.82% 84.83%
99 Qwen3 235B A22B Instruct 2507 40.15% 83.15% 80.10%
100 Claude 3 Haiku 39.32% 68.47% 71.19%
101 Gemini 2.5 Flash 38.47% 61.45% 80.60%
102 Z.AI GLM 4.5 38.11% 79.19% 86.27%
103 Ministral 3 8B 38.01% 74.43% 71.76%
104 Ministral 3 3B 37.04% 72.38% 67.22%
105 Writer: Palmyra X5 35.61% 79.71% 79.57%
106 Cohere Command R+ (Aug. 2024) 35.13% 59.51% 69.03%
107 Mistral Small Creative 33.57% 76.28% 73.27%
108 Qwen 3 32B 31.96% 81.66% 82.21%
109 Mistral Large 31.90% 73.04% 80.15%
110 Hermes 3 70B 28.43% 61.15% 72.57%
111 Arcee AI: Trinity Mini 26.25% 59.94% 70.90%
112 Gemma 3 4B 22.12% 60.30% 68.57%
113 Ministral 8B 21.93% 46.82% 64.87%
114 Mistral Large 2 21.18% 69.19% 82.41%
115 WizardLM 2 8x22b 20.73% 67.14% 71.07%
116 Ministral 3B 17.87% 49.17% 61.29%
117 Mistral NeMO 15.70% 51.55% 65.04%
118 Rocinante 12B 14.07% 48.47% 54.55%