Comprehension
Subcategory of Language. 155 models scored.
Model Leaderboard
All models ranked by their Comprehension subcategory score.
| # | Model | Comprehension | Language | Overall |
|---|---|---|---|---|
| 1 | Qwen3.7 Max | 100.00% | 97.05% | 95.75% |
| 2 | Claude Opus 4.6 (Reasoning) | 100.00% | 96.12% | 95.02% |
| 3 | Qwen3.6 Max Preview | 100.00% | 100.00% | 94.54% |
| 4 | Z.AI GLM 5 Turbo | 100.00% | 99.90% | 94.27% |
| 5 | Claude Sonnet 4.6 (Reasoning) | 100.00% | 97.58% | 93.66% |
| 6 | Claude Opus 4.7 (Reasoning) | 100.00% | 98.77% | 93.23% |
| 7 | GPT-5.5 (Reasoning) | 100.00% | 99.69% | 92.98% |
| 8 | GPT-5.5 (Reasoning, Low) | 100.00% | 99.24% | 92.59% |
| 9 | Claude Opus 4.6 | 100.00% | 96.13% | 92.35% |
| 10 | MoonshotAI: Kimi K2.6 | 100.00% | 96.77% | 92.31% |
| 11 | Qwen 3.5 397B A17B | 100.00% | 95.01% | 91.73% |
| 12 | Qwen 3.5 122B | 100.00% | 95.01% | 91.53% |
| 13 | Qwen 3.5 Plus (2026-04-20) | 100.00% | 97.14% | 91.51% |
| 14 | Grok 4.20 (Beta, Reasoning) | 100.00% | 99.08% | 91.49% |
| 15 | Claude Sonnet 4.6 | 100.00% | 100.00% | 91.15% |
| 16 | MoonshotAI: Kimi K2.5 | 100.00% | 97.10% | 91.04% |
| 17 | MiniMax M3 | 100.00% | 94.71% | 90.88% |
| 18 | Qwen 3.5 27B | 100.00% | 95.53% | 90.85% |
| 19 | ByteDance Seed 1.6 | 100.00% | 95.63% | 90.70% |
| 20 | GPT-5.4 Mini (Reasoning) | 100.00% | 98.12% | 90.65% |
| 21 | Claude Opus 4.5 | 100.00% | 99.66% | 89.69% |
| 22 | Aion 2.0 | 100.00% | 96.17% | 89.21% |
| 23 | Z.AI GLM 4.6 | 100.00% | 96.60% | 89.11% |
| 24 | MiniMax M2.7 | 100.00% | 84.80% | 89.10% |
| 25 | DeepSeek V4 Flash (Reasoning) | 100.00% | 94.76% | 89.01% |
| 26 | Qwen 3.5 35B | 100.00% | 91.95% | 88.00% |
| 27 | Claude Opus 4 | 100.00% | 93.01% | 87.69% |
| 28 | Qwen 3.5 Plus (2026-02-15) | 100.00% | 95.10% | 85.96% |
| 29 | Mistral Large 3 | 100.00% | 92.02% | 85.43% |
| 30 | GPT-4o, May 13th (temp=0) | 100.00% | 98.72% | 85.36% |
| 31 | DeepSeek-V2 Chat | 100.00% | 100.00% | 84.83% |
| 32 | ByteDance Seed 2.0 Lite | 100.00% | 96.80% | 84.80% |
| 33 | DeepSeek V3 (2024-12-26) | 100.00% | 87.88% | 83.68% |
| 34 | Hermes 3 405B | 100.00% | 99.57% | 82.86% |
| 35 | Mistral Large 2 | 100.00% | 85.22% | 82.41% |
| 36 | DeepSeek V3 (2025-03-24) | 100.00% | 86.42% | 81.99% |
| 37 | Ministral 3 3B | 100.00% | 68.10% | 67.22% |
| 38 | Gemini 3.1 Pro (Preview) | 95.00% | 94.90% | 94.37% |
| 39 | Gemini 3.5 Flash (Reasoning) | 95.00% | 94.41% | 94.08% |
| 40 | Grok 4.3 (Reasoning) | 95.00% | 97.50% | 93.60% |
| 41 | GPT-5.4 (Reasoning) | 95.00% | 94.90% | 93.24% |
| 42 | GPT-5 Mini | 95.00% | 96.49% | 92.62% |
| 43 | Claude Opus 4.8 (Reasoning) | 95.00% | 96.38% | 92.22% |
| 44 | Claude Opus 4.8 (Reasoning, Low) | 95.00% | 96.31% | 92.14% |
| 45 | Gemma 4 26B (Reasoning) | 95.00% | 95.20% | 91.49% |
| 46 | Grok 4.20 (Reasoning) | 95.00% | 96.61% | 91.39% |
| 47 | Z.AI GLM 5 | 95.00% | 92.06% | 91.23% |
| 48 | Gemini 3 Flash (Preview, Reasoning) | 95.00% | 94.93% | 90.50% |
| 49 | Gemini 3 Pro (Preview) | 95.00% | 89.64% | 88.79% |
| 50 | MiniMax M2.5 | 95.00% | 96.05% | 88.71% |
| 51 | Claude Sonnet 4.5 | 95.00% | 92.39% | 88.03% |
| 52 | Stealth: Hunter Alpha | 95.00% | 93.35% | 87.34% |
| 53 | Gemini 3.5 Flash (Reasoning, Minimal) | 95.00% | 97.50% | 86.47% |
| 54 | GPT-OSS 120B | 95.00% | 97.18% | 86.44% |
| 55 | Gemini 3.1 Flash Lite (Reasoning) | 95.00% | 96.70% | 86.41% |
| 56 | Qwen 3.5 Flash | 95.00% | 91.94% | 86.38% |
| 57 | Z.AI GLM 4.5 | 95.00% | 97.33% | 86.27% |
| 58 | Gemini 3.1 Flash Lite (Preview) | 95.00% | 94.98% | 85.87% |
| 59 | Claude Haiku 4.5 | 95.00% | 91.84% | 85.14% |
| 60 | Z.AI GLM 4.7 Flash | 95.00% | 87.67% | 84.82% |
| 61 | GPT-4o, May 13th (temp=1) | 95.00% | 92.52% | 83.80% |
| 62 | Z.AI GLM 4.5 Air | 95.00% | 95.05% | 83.12% |
| 63 | DeepSeek V3.1 | 95.00% | 96.87% | 82.39% |
| 64 | Cydonia 24B V4.1 | 95.00% | 72.49% | 75.09% |
| 65 | WizardLM 2 8x22b | 95.00% | 78.05% | 71.06% |
| 66 | Mistral NeMO | 95.00% | 80.80% | 65.04% |
| 67 | GPT-5.1 | 90.00% | 93.64% | 92.54% |
| 68 | Qwen 3.6 27B | 90.00% | 89.01% | 89.72% |
| 69 | Grok 4.1 Fast | 90.00% | 88.76% | 89.55% |
| 70 | GPT-5.5 | 90.00% | 94.15% | 89.09% |
| 71 | Qwen 3.6 35B | 90.00% | 93.56% | 89.05% |
| 72 | Claude Sonnet 4 | 90.00% | 91.31% | 88.72% |
| 73 | GPT-4.1 | 90.00% | 93.91% | 88.68% |
| 74 | Gemini 2.5 Pro | 90.00% | 92.57% | 88.53% |
| 75 | ByteDance Seed 2.0 Mini | 90.00% | 90.12% | 86.91% |
| 76 | Qwen 3.5 9B | 90.00% | 88.18% | 86.05% |
| 77 | Stealth: Healer Alpha | 90.00% | 88.45% | 85.93% |
| 78 | Gemma 4 26B | 90.00% | 92.02% | 85.84% |
| 79 | Gemini 3 Flash (Preview) | 90.00% | 95.00% | 85.35% |
| 80 | Claude 3.7 Sonnet | 90.00% | 92.95% | 83.39% |
| 81 | Qwen 3 32B | 90.00% | 84.61% | 82.21% |
| 82 | DeepSeek V4 Flash | 90.00% | 88.50% | 82.02% |
| 83 | Mistral Large | 90.00% | 88.64% | 80.15% |
| 84 | Z.AI GLM 5.1 | 85.00% | 91.57% | 94.37% |
| 85 | GPT-5 | 85.00% | 91.50% | 91.93% |
| 86 | GPT-5.2 | 85.00% | 91.19% | 90.26% |
| 87 | Xiaomi MIMO v2.5 Pro | 85.00% | 87.69% | 87.36% |
| 88 | Gemini 3.1 Flash Lite | 85.00% | 90.90% | 85.75% |
| 89 | Grok 4.20 (Beta) | 85.00% | 91.17% | 83.85% |
| 90 | GPT-5.4 (Reasoning, Low) | 85.00% | 90.79% | 91.41% |
| 91 | Qwen 3.6 Flash | 85.00% | 89.33% | 90.65% |
| 92 | DeepSeek V4 Pro (Reasoning) | 85.00% | 88.51% | 90.10% |
| 93 | Claude Opus 4.7 | 85.00% | 92.32% | 89.93% |
| 94 | Grok 4 | 85.00% | 90.61% | 88.12% |
| 95 | Grok 4 Fast | 85.00% | 84.61% | 86.15% |
| 96 | GPT-5.4 Mini (Reasoning, Low) | 85.00% | 92.45% | 85.75% |
| 97 | Stealth: Aurora Alpha | 85.00% | 92.50% | 83.79% |
| 98 | Z.AI GLM 4.7 | 80.00% | 85.46% | 88.69% |
| 99 | GPT-4.1 Mini | 80.00% | 89.64% | 83.20% |
| 100 | GPT-5.4 Mini | 80.00% | 88.75% | 82.43% |
| 101 | DeepSeek V3.2 | 80.00% | 85.01% | 82.25% |
| 102 | Gemini 2.5 Flash Lite | 80.00% | 82.75% | 81.08% |
| 103 | Nemotron 3 Nano | 80.00% | 87.63% | 77.73% |
| 104 | Arcee AI: Trinity Large (Preview) | 80.00% | 78.38% | 73.33% |
| 105 | Skyfall 36B V2 | 80.00% | 73.94% | 65.76% |
| 106 | Gemma 4 31B (Reasoning) | 75.00% | 83.82% | 91.71% |
| 107 | Gemini 2.5 Flash (Reasoning) | 75.00% | 86.06% | 86.51% |
| 108 | Claude 3.5 Sonnet | 75.00% | 85.62% | 84.24% |
| 109 | Inception Mercury 2 | 75.00% | 87.32% | 83.85% |
| 110 | Gemini 2.5 Flash | 75.00% | 86.23% | 80.60% |
| 111 | Qwen3 235B A22B Instruct 2507 | 75.00% | 60.83% | 80.10% |
| 112 | Grok 4.3 | 75.00% | 84.74% | 78.66% |
| 113 | Mistral Small 3.2 24B | 75.00% | 72.77% | 78.58% |
| 114 | Llama 3.1 70B | 75.00% | 80.18% | 78.40% |
| 115 | Hermes 3 70B | 75.00% | 81.66% | 72.57% |
| 116 | Rocinante 12B | 75.00% | 63.45% | 54.54% |
| 117 | DeepSeek V4 Pro | 70.00% | 72.80% | 82.63% |
| 118 | GPT-5.4 Nano (Reasoning) | 70.00% | 83.99% | 81.36% |
| 119 | Writer: Palmyra X5 | 70.00% | 56.58% | 79.57% |
| 120 | GPT-5.4 Nano (Reasoning, Low) | 70.00% | 81.87% | 79.48% |
| 121 | Gemma 3 12B | 70.00% | 80.10% | 78.41% |
| 122 | Qwen 2.5 72B | 70.00% | 68.95% | 75.46% |
| 123 | GPT-5.4 Nano | 70.00% | 80.82% | 74.40% |
| 124 | Gemma 3 4B | 70.00% | 72.28% | 68.57% |
| 125 | Gemini 2.5 Flash Lite (Reasoning) | 65.00% | 74.36% | 85.75% |
| 126 | Nemotron 3 Super | 65.00% | 81.41% | 84.56% |
| 127 | GPT-5.4 | 65.00% | 81.49% | 84.32% |
| 128 | GPT-4o, Aug. 6th (temp=1) | 65.00% | 82.21% | 82.62% |
| 129 | Grok 4.20 | 65.00% | 78.86% | 81.70% |
| 130 | Inception Mercury | 65.00% | 80.37% | 79.50% |
| 131 | Gemma 3 27B | 65.00% | 77.21% | 77.85% |
| 132 | GPT-4.1 Nano | 65.00% | 78.95% | 71.94% |
| 133 | LFM2 24B | 65.00% | 64.64% | 58.77% |
| 134 | Xiaomi MIMO v2.5 | 65.00% | 76.33% | 85.05% |
| 135 | Claude 3 Haiku | 65.00% | 72.76% | 71.19% |
| 136 | Llama 3.1 Nemotron 70B | 60.00% | 46.80% | 74.70% |
| 137 | Arcee AI: Trinity Mini | 60.00% | 70.59% | 70.90% |
| 138 | o4 Mini High | 60.00% | 79.76% | 90.29% |
| 139 | o4 Mini | 60.00% | 80.00% | 88.35% |
| 140 | Cohere Command R+ (Aug. 2024) | 60.00% | 66.58% | 69.03% |
| 141 | GPT-5 Nano | 55.00% | 77.18% | 82.60% |
| 142 | GPT-4o Mini (temp=1) | 55.00% | 77.50% | 79.08% |
| 143 | Mistral Small 4 | 55.00% | 51.96% | 76.46% |
| 144 | Ministral 8B | 55.00% | 53.91% | 64.87% |
| 145 | Llama 3.1 8B | 55.00% | 64.06% | 63.35% |
| 146 | Gemma 4 31B | 50.00% | 75.00% | 86.91% |
| 147 | GPT-4o, Aug. 6th (temp=0) | 50.00% | 75.00% | 82.45% |
| 148 | Mistral Small 4 (Reasoning) | 50.00% | 60.53% | 82.39% |
| 149 | GPT-4o Mini (temp=0) | 50.00% | 75.00% | 78.29% |
| 150 | Mistral Medium 3.1 | 50.00% | 49.50% | 77.83% |
| 151 | Mistral Small Creative | 50.00% | 41.85% | 73.27% |
| 152 | Ministral 3 14B | 50.00% | 30.00% | 72.54% |
| 153 | Ministral 3 8B | 50.00% | 48.96% | 71.76% |
| 154 | ByteDance Seed 1.6 Flash | 40.00% | 61.23% | 73.27% |
| 155 | Ministral 3B | 25.00% | 42.25% | 61.29% |