Transformation

Subcategory of Text Editing. 89 models scored.

Model Leaderboard

All models ranked by their Transformation subcategory score.

# Model Transformation Text Editing Overall
1 Gemini 3 Pro (Preview) 98.12% 98.86% 88.79%
2 GPT-5 98.11% 98.90% 91.93%
3 Claude Opus 4.6 (Reasoning) 97.85% 98.86% 95.02%
4 Qwen 3.5 397B A17B 97.78% 98.05% 91.73%
5 Claude Sonnet 4.5 97.74% 99.02% 88.03%
6 Claude Sonnet 4 97.70% 99.13% 88.72%
7 ByteDance Seed 1.6 97.62% 98.40% 90.70%
8 Qwen 3.5 27B 97.53% 98.69% 90.85%
9 GPT-5.1 97.43% 98.54% 92.54%
10 Z.AI GLM 5 97.15% 98.59% 91.23%
11 Grok 4 97.06% 98.76% 88.12%
12 Gemini 2.5 Flash (Reasoning) 96.95% 98.12% 86.51%
13 Gemini 2.5 Pro 96.93% 98.58% 88.53%
14 Gemini 3.1 Pro (Preview) 96.51% 98.51% 94.37%
15 MoonshotAI: Kimi K2.5 96.37% 97.79% 91.04%
16 Grok 4.1 Fast 96.32% 97.87% 89.55%
17 GPT-5.2 96.25% 97.54% 90.26%
18 Gemini 3 Flash (Preview, Reasoning) 96.17% 98.12% 90.50%
19 Z.AI GLM 4.7 95.98% 98.22% 88.69%
20 Claude Sonnet 4.6 (Reasoning) 95.84% 98.30% 93.66%
21 DeepSeek V3.2 95.62% 95.78% 82.25%
22 Z.AI GLM 4.5 95.45% 95.32% 86.27%
23 Claude Opus 4.6 95.06% 98.35% 92.35%
24 Z.AI GLM 4.6 95.02% 97.78% 89.11%
25 Qwen 3.5 Plus (2026-02-15) 94.91% 98.10% 85.96%
26 Gemini 2.5 Flash 94.82% 97.83% 80.60%
27 Gemini 2.5 Flash Lite (Reasoning) 94.57% 94.54% 85.75%
28 GPT-5 Mini 93.87% 97.13% 92.62%
29 Gemini 3 Flash (Preview) 93.70% 97.54% 85.35%
30 Claude 3.7 Sonnet 93.64% 97.12% 83.39%
31 Aion 2.0 93.58% 95.34% 89.21%
32 Claude Opus 4 93.42% 97.25% 87.69%
33 Grok 4 Fast 93.19% 97.26% 86.15%
34 Claude Opus 4.5 93.08% 97.69% 89.69%
35 Claude 3.5 Sonnet 92.76% 96.57% 84.24%
36 Qwen 3.5 122B 91.84% 96.31% 91.53%
37 Qwen 3.5 35B 91.82% 94.95% 88.00%
38 Mistral Large 2 90.87% 94.16% 82.41%
39 Claude Haiku 4.5 90.77% 96.81% 85.14%
40 Mistral Large 90.77% 95.14% 80.15%
41 Mistral Large 3 90.68% 94.09% 85.43%
42 Minimax M2.5 90.05% 96.02% 88.71%
43 DeepSeek V3 (2024-12-26) 89.33% 93.58% 83.68%
44 Claude Sonnet 4.6 89.29% 96.37% 91.15%
45 GPT-4.1 Mini 89.26% 95.62% 83.20%
46 Qwen 3.5 Flash 88.93% 92.80% 86.38%
47 GPT-4o, Aug. 6th (temp=0) 88.11% 93.77% 82.45%
48 GPT-4o, May 13th (temp=0) 87.89% 95.35% 85.36%
49 o4 Mini High 87.18% 94.36% 90.29%
50 GPT-4.1 86.53% 94.40% 88.68%
51 Mistral Medium 3.1 84.76% 93.77% 77.83%
52 Gemini 2.5 Flash Lite 84.21% 92.13% 81.08%
53 Llama 3.1 70B 84.01% 92.10% 78.40%
54 DeepSeek-V2 Chat 83.56% 90.90% 84.83%
55 ByteDance Seed 1.6 Flash 81.97% 91.64% 73.27%
56 DeepSeek V3.1 81.88% 87.27% 82.39%
57 WizardLM 2 8x22b 81.62% 88.13% 71.07%
58 GPT-4o, May 13th (temp=1) 79.68% 92.41% 83.80%
59 DeepSeek V3 (2025-03-24) 79.33% 89.57% 81.99%
60 Writer: Palmyra X5 79.31% 91.20% 79.57%
61 o4 Mini 77.98% 90.61% 88.35%
62 Arcee AI: Trinity Large (Preview) 76.70% 86.62% 73.33%
63 Mistral Small 3.2 24B 76.07% 89.48% 78.60%
64 Mistral Small Creative 73.96% 90.31% 73.27%
65 Llama 3.1 Nemotron 70B 73.30% 87.26% 74.70%
66 GPT-4o, Aug. 6th (temp=1) 72.28% 86.72% 82.62%
67 Qwen 2.5 72B 71.74% 89.18% 75.46%
68 Z.AI GLM 4.7 Flash 70.63% 85.82% 84.82%
69 GPT-4o Mini (temp=1) 70.40% 85.78% 79.08%
70 Hermes 3 405B 69.94% 89.14% 82.86%
71 Ministral 3 14B 69.48% 86.20% 72.54%
72 Gemma 3 12B 67.87% 85.18% 78.41%
73 GPT-4o Mini (temp=0) 67.00% 84.62% 78.29%
74 LFM2 24B 66.33% 71.56% 58.77%
75 Gemma 3 27B 64.41% 86.63% 77.85%
76 GPT-5 Nano 61.06% 82.74% 82.60%
77 Ministral 3 8B 52.05% 78.52% 71.76%
78 Llama 3.1 8B 51.53% 75.45% 63.37%
79 Arcee AI: Trinity Mini 50.86% 73.88% 70.90%
80 GPT-4.1 Nano 50.14% 76.06% 71.94%
81 Ministral 8B 48.88% 77.52% 64.87%
82 Claude 3 Haiku 48.58% 64.36% 71.19%
83 Mistral NeMO 45.51% 73.69% 65.04%
84 Hermes 3 70B 44.57% 63.34% 72.57%
85 Gemma 3 4B 40.80% 78.38% 68.57%
86 Ministral 3B 40.57% 70.91% 61.29%
87 Ministral 3 3B 39.84% 69.80% 67.22%
88 Cohere Command R+ (Aug. 2024) 39.15% 68.40% 69.03%
89 Rocinante 12B 32.19% 56.31% 54.55%