HyperAI
Command Palette
Search for a command to run...
Humanity's Last Exam
수학, 과학, 인문학에 걸친 전문가 검증 2,500문항으로 구성된 프런티어 수준 벤치마크로, 최종적인 폐쇄형 학술 평가를 목표로 설계되었습니다.
- 모델
- 267 현재, 684 전체
- 중앙값
- 12.0%
- 범위
- 1.1%–74.3%
| 순위 | 비교 | 모델 | 점수 |
|---|---|---|---|
| 201 | 5.5% | ||
| 202 |  Apertus 70B Instruct Swiss AI Initiative | 5.5% | |
| 203 | 5.5% | ||
| 204 |  K2 Horizon 0.9B Institute of Foundation Models | 5.4% | |
| 205 |  Ministral 3 3B Mistral | 5.4% | |
| 206 |  gpt-oss-20b (Low) OpenAI | 5.3% | |
| 207 |  Tiny Aya Global Cohere | 5.2% | |
| 208 | 5.2% | ||
| 209 |  Molmo 7B-D Allen Institute for AI | 5.1% | |
| 210 |  LFM2.5-VL-1.6B Liquid AI | 5.1% | |
| 211 |  Qwen3.5 0.8B (Non-reasoning) Alibaba | 5.1% | |
| 212 |  MiniCPM-V 4.6 1.3B OpenBMB | 5.1% | |
| 213 |  EXAONE 4.0 32B (Non-reasoning) LG AI Research | 5.0% | |
| 214 | 5.0% | ||
| 215 | 5.0% | ||
| 216 | Olmo 3.1 32B Instruct Allen Institute for AI | 5.0% | |
| 217 |  Phi-4 Multimodal Instruct Microsoft | 5.0% | |
| 218 | Apertus 8B Instruct Swiss AI Initiative | 5.0% | |
| 219 | Qwen3.5 2B (Non-reasoning) Alibaba | 5.0% | |
| 220 |  Llama 4 Maverick Meta | 4.9% | |
| 221 | 4.9% | ||
| 222 | 4.9% | ||
| 223 | 4.8% | ||
| 224 | 4.8% | ||
| 225 |  Gemma 4 E2B (Reasoning) Google | 4.8% | |
| 226 | 4.7% | ||
| 227 | Ministral 3 14B Mistral | 4.6% | |
| 228 |  Nova Micro Amazon | 4.6% | |
| 229 | Qwen3 Omni 30B A3B Instruct Alibaba | 4.6% | |
| 230 | 4.6% | ||
| 231 | 4.6% | ||
| 232 | MiniCPM5-1B (Non-reasoning) OpenBMB | 4.5% | |
| 233 |  G9v3-3B AI9Stars | 4.5% | |
| 234 | 4.5% | ||
| 235 | Phi-4 Mini Instruct Microsoft | 4.5% | |
| 236 |  Jamba 1.7 Mini AI21 Labs | 4.4% | |
| 237 |  Reka Flash 3 Reka AI | 4.4% | |
| 238 | K2-V2 (Medium) Institute of Foundation Models | 4.4% | |
| 239 | Ministral 3 8B Mistral | 4.3% | |
| 240 | Molmo2-8B Allen Institute for AI | 4.3% | |
| 241 | 4.3% | ||
| 242 |  DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning) Nous Research | 4.3% | |
| 243 | Hermes 4 - Llama-3.1 405B (Non-reasoning) Nous Research | 4.2% | |
| 244 | LFM2 24B A2B Liquid AI | 4.2% | |
| 245 | Nova Premier Amazon | 4.2% | |
| 246 | Command A Cohere | 4.0% | |
| 247 | Nova 2.0 Lite (Low) Amazon | 4.0% | |
| 248 | 4.0% | ||
| 249 | 3.9% | ||
| 250 | 3.8% |
