HyperAI
Command Palette
Search for a command to run...
LiveCodeBench
LeetCode, AtCoder, CodeForces에서 새로운 경쟁 프로그래밍 문제를 지속적으로 수집하는 오염 없는 코딩 벤치마크로, 코드 생성, 자기 수정, 실행을 평가합니다.
- 모델
- 82 현재, 343 전체
- 중앙값
- 47.2%
- 범위
- 0.3%–87.8%
| 순위 | 비교 | 모델 | 점수 |
|---|---|---|---|
| 1 |  gpt-oss-120b (High) OpenAI | 87.8% | |
| 2 | 81.2% | ||
| 3 | o3 OpenAI | 80.8% | |
| 4 |  Apriel-v1.6-15B-Thinker ServiceNow | 80.7% | |
| 5 | 78.4% | ||
| 6 | gpt-oss-20b (High) OpenAI | 77.7% | |
| 7 |  INTELLECT-3 (Based on GLM-4.5-Air) Prime Intellect | 77.7% | |
| 8 |  Doubao Seed Code ByteDance Seed | 76.6% | |
| 9 |  Magistral Medium 1.2 Mistral | 75.0% | |
| 10 |  EXAONE 4.0 32B (Reasoning) LG AI Research | 74.7% | |
| 11 | 74.1% | ||
| 12 | 73.0% | ||
| 13 |  Falcon-H1R-7B TII UAE | 72.4% | |
| 14 | 72.4% | ||
| 15 | Magistral Small 1.2 Mistral | 72.3% | |
| 16 |  Nova 2.0 Lite (High) Amazon | 71.1% | |
| 17 | gpt-oss-120b (Low) OpenAI | 70.7% | |
| 18 | 70.1% | ||
| 19 |  Olmo 3.1 32B Think Allen Institute for AI | 69.5% | |
| 20 |  K2-V2 (High) Institute of Foundation Models | 69.4% | |
| 21 | 69.4% | ||
| 22 |  Cogito v2.1 (Reasoning) Deep Cogito | 68.8% | |
| 23 |  Hermes 4 - Llama-3.1 405B (Reasoning) Nous Research | 68.6% | |
| 24 |  Qwen3 Next 80B A3B Instruct Alibaba | 68.4% | |
| 25 | 67.9% | ||
| 26 | Nova 2.0 Lite (Medium) Amazon | 66.3% | |
| 27 | Nova 2.0 Omni (Medium) Amazon | 66.0% | |
| 28 |  Mi:dm K 2.5 Pro Korea Telecom | 65.6% | |
| 29 | Hermes 4 - Llama-3.1 70B (Reasoning) Nous Research | 65.3% | |
| 30 | gpt-oss-20b (Low) OpenAI | 65.2% | |
| 31 |  Motif-2-12.7B-Reasoning Motif Technologies | 65.1% | |
| 32 | 64.1% | ||
| 33 | 63.8% | ||
| 34 | 62.9% | ||
| 35 |  Ring-flash-2.0 InclusionAI | 62.8% | |
| 36 | Olmo 3 7B Think Allen Institute for AI | 61.7% | |
| 37 | Nova 2.0 Omni (Low) Amazon | 59.2% | |
| 38 | Hermes 4 - Llama-3.1 405B (Non-reasoning) Nous Research | 54.6% | |
| 39 | K2-V2 (Medium) Institute of Foundation Models | 54.1% | |
| 40 | Exaone 4.0 1.2B (Reasoning) LG AI Research | 51.6% | |
| 41 | 47.3% | ||
| 42 | EXAONE 4.0 32B (Non-reasoning) LG AI Research | 47.2% | |
| 43 | Nova 2.0 Lite (Low) Amazon | 46.9% | |
| 44 |  ERNIE 4.5 300B A47B Baidu | 46.7% | |
| 45 |  Reka Flash 3 Reka AI | 43.5% | |
| 46 | Qwen3 Omni 30B A3B Instruct Alibaba | 42.2% | |
| 47 |  Llama 4 Maverick Meta | 39.7% | |
| 48 | K2-V2 (Low) Institute of Foundation Models | 39.3% | |
| 49 | 37.8% | ||
| 50 | 36.0% |
