HyperAI
Command Palette
Search for a command to run...
AA-LCR v1.1: Long Context Reasoning
10k에서 100k 토큰에 이르는 장문 문서에서 정보를 추출하고 추론하며 종합하는 언어 모델의 능력을 측정하는 어려운 벤치마크입니다(cl100k_base 토크나이저로 측정).
- 모델
- 259 현재, 596 전체
- 중앙값
- 61.3%
- 범위
- 0.0%–88.3%
| 순위 | 비교 | 모델 | 점수 |
|---|---|---|---|
| 101 |  Mercury 2.5 Inception | 71.7% | |
| 102 |  Qwen3.6 35B A3B (Reasoning) Alibaba | 71.7% | |
| 103 |  Solar Open2 250B Upstage | 71.7% | |
| 104 |  JT-4.1 Flash 236B A21B (Non-reasoning) China Mobile | 71.3% | |
| 105 | Solar Pro 4 Upstage | 70.7% | |
| 106 |  A.X-K2 SK Telecom | 70.0% | |
| 107 | 70.0% | ||
| 108 | Qwen3.5 9B (Reasoning) Alibaba | 70.0% | |
| 109 |  Ring-2.6-1T InclusionAI | 70.0% | |
| 110 |  Gemma 4 31B (Reasoning) Google | 69.7% | |
| 111 |  K2 Horizon 7B Institute of Foundation Models | 69.7% | |
| 112 | 69.7% | ||
| 113 |  Mistral Medium 3.5 Mistral | 69.3% | |
| 114 | Qwen3.8 27B (Non-reasoning) Alibaba | 69.3% | |
| 115 |  Nemotron 3.5 Super VL NVIDIA | 69.0% | |
| 116 | 67.7% | ||
| 117 | 65.7% | ||
| 118 | JT-35B-Flash China Mobile | 65.7% | |
| 119 |  G9v3-39A5B AI9Stars | 65.3% | |
| 120 |  LongCat 2.0 LongCat | 65.0% | |
| 121 | 64.7% | ||
| 122 | 64.3% | ||
| 123 | 64.3% | ||
| 124 | 64.0% | ||
| 125 | Gemma 4 12B (Reasoning) Google | 63.7% | |
| 126 | Qwen3.5 Omni Plus Alibaba | 63.7% | |
| 127 | 63.7% | ||
| 128 | Qwen3.5 4B (Reasoning) Alibaba | 63.0% | |
| 129 | K2 Horizon 3.7B Institute of Foundation Models | 62.3% | |
| 130 | 61.3% | ||
| 131 | Ling 3.0 Tiny InclusionAI | 60.3% | |
| 132 | Nemotron 3.5 Lightning NVIDIA | 60.3% | |
| 133 |  Nova 2.0 Lite (High) Amazon | 60.3% | |
| 134 |  K-EXAONE 2.0 LG AI Research | 60.0% | |
| 135 | Nova 2.0 Lite (Medium) Amazon | 60.0% | |
| 136 | Nova 2.0 Omni (Medium) Amazon | 59.7% | |
| 137 |  MiniCPM5-2B OpenBMB | 59.3% | |
| 138 | Nova 2.0 Omni (Low) Amazon | 57.7% | |
| 139 | 57.3% | ||
| 140 | K2 Think V2 Institute of Foundation Models | 57.0% | |
| 141 | EXAONE 4.5 33B (Reasoning) LG AI Research | 54.7% | |
| 142 | Nova 2.0 Lite (Low) Amazon | 54.3% | |
| 143 | Magistral Medium 1.2 Mistral | 53.0% | |
| 144 |  Command A+ Cohere | 52.7% | |
| 145 | Qwen3 Next 80B A3B Instruct Alibaba | 52.7% | |
| 146 |  gpt-oss-120b (High) OpenAI | 52.0% | |
| 147 | Qwen3.5 Omni Flash Alibaba | 52.0% | |
| 148 | 50.7% | ||
| 149 |  Llama 4 Maverick Meta | 50.0% | |
| 150 | Mistral Small 4 (Reasoning) Mistral | 49.7% |
