벤치마크 상세
SciCode
과학 분야 코드 작성: 과학 분야 코드 작성
공식 리더보드 보기
과학 연구용 코드 생성의 공개 평가 결과입니다. 일반 저장소 버그 수정이나 앱 구현 점수와 동일하게 해석하지 않습니다.
모든 공개 결과
모델 검색 · 점수순확인 2026. 09. 07.SciCode v1.0.1 · Artificial Analysis% correct
107개 결과원문 ↗
| 순서 | 모델 / 실행 구성 | 점수 | 평가 조건·근거 |
|---|---|---|---|
| 1 | Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)Anthropic | 63.08 | SciCode v1.0.1 · Artificial Analysis |
| 2 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)Anthropic | 61 | SciCode v1.0.1 · Artificial Analysis |
| 3 | Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)Anthropic | 60.88 | SciCode v1.0.1 · Artificial Analysis |
| 4 | Gemini 3.7 Flash (medium)Google | 59.84 | SciCode v1.0.1 · Artificial Analysis |
| 5 | Muse Spark 1.3 (xhigh)Meta | 59.72 | SciCode v1.0.1 · Artificial Analysis |
| 6 | Kimi K3 (max)Kimi | 59.49 | SciCode v1.0.1 · Artificial Analysis |
| 7 | GLM-5.3 (max)Z AI | 59.03 | SciCode v1.0.1 · Artificial Analysis |
| 8 | Muse Spark 1.1 (xhigh)Meta | 58.8 | SciCode v1.0.1 · Artificial Analysis |
| 9 | Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)Anthropic | 58.68 | SciCode v1.0.1 · Artificial Analysis |
| 10 | Gemini 3.1 Pro PreviewGoogle | 58.68 | SciCode v1.0.1 · Artificial Analysis |
| 11 | Muse Spark 1.3 (max)Meta | 58.33 | SciCode v1.0.1 · Artificial Analysis |
| 12 | GPT-5.6 Sol (high)OpenAI | 57.75 | SciCode v1.0.1 · Artificial Analysis |
| 13 | Muse Spark 1.2 (xhigh)Meta | 57.41 | SciCode v1.0.1 · Artificial Analysis |
| 14 | GPT-5.6 Sol (medium)OpenAI | 57.41 | SciCode v1.0.1 · Artificial Analysis |
| 15 | Gemini 3.7 Flash (high)Google | 57.18 | SciCode v1.0.1 · Artificial Analysis |
| 16 | GPT-5.6 Sol (max)OpenAI | 57.06 | SciCode v1.0.1 · Artificial Analysis |
| 17 | GPT-5.6 Sol (xhigh)OpenAI | 57.06 | SciCode v1.0.1 · Artificial Analysis |
| 18 | Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback)Anthropic | 56.71 | SciCode v1.0.1 · Artificial Analysis |
| 19 | Gemini 3.8 Flash (high)Google | 56.6 | SciCode v1.0.1 · Artificial Analysis |
| 20 | GPT-6 Astra (max)OpenAI | 56.48 | SciCode v1.0.1 · Artificial Analysis |
| 21 | Grok 4.6 (high)SpaceXAI | 56.48 | SciCode v1.0.1 · Artificial Analysis |
| 22 | GPT-5.6 Sol (low)OpenAI | 56.37 | SciCode v1.0.1 · Artificial Analysis |
| 23 | Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)Anthropic | 56.37 | SciCode v1.0.1 · Artificial Analysis |
| 24 | Claude Opus 5 (Adaptive Reasoning, Max Effort)Anthropic | 56.37 | SciCode v1.0.1 · Artificial Analysis |
| 25 | Grok 4.6 (medium)SpaceXAI | 55.9 | SciCode v1.0.1 · Artificial Analysis |
| 26 | Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)Anthropic | 55.67 | SciCode v1.0.1 · Artificial Analysis |
| 27 | GPT-6 Astra (xhigh)OpenAI | 55.67 | SciCode v1.0.1 · Artificial Analysis |
| 28 | Gemini 3.7 Flash (low)Google | 55.67 | SciCode v1.0.1 · Artificial Analysis |
| 29 | Claude Opus 5 (Adaptive Reasoning, High Effort)Anthropic | 55.44 | SciCode v1.0.1 · Artificial Analysis |
| 30 | GPT-6 Astra (high)OpenAI | 55.44 | SciCode v1.0.1 · Artificial Analysis |
| 31 | Gemini 3.8 Flash (medium)Google | 55.09 | SciCode v1.0.1 · Artificial Analysis |
| 32 | Grok 4.5 (high)SpaceXAI | 54.98 | SciCode v1.0.1 · Artificial Analysis |
| 33 | GPT-5.6 Terra (max)OpenAI | 54.98 | SciCode v1.0.1 · Artificial Analysis |
| 34 | Gemini 3.8 Flash (low)Google | 54.98 | SciCode v1.0.1 · Artificial Analysis |
| 35 | GPT-5.5 (medium)OpenAI | 54.51 | SciCode v1.0.1 · Artificial Analysis |
| 36 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort)Anthropic | 54.4 | SciCode v1.0.1 · Artificial Analysis |
| 37 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort)Anthropic | 54.28 | SciCode v1.0.1 · Artificial Analysis |
| 38 | GPT-6 Astra (medium)OpenAI | 54.17 | SciCode v1.0.1 · Artificial Analysis |
| 39 | Qwen3.8 2.4T A95BAlibaba | 54.05 | SciCode v1.0.1 · Artificial Analysis |
| 40 | GPT-6 Astra (low)OpenAI | 54.05 | SciCode v1.0.1 · Artificial Analysis |
| 41 | Gemini 3.5 Flash (high)Google | 53.94 | SciCode v1.0.1 · Artificial Analysis |
| 42 | GPT-5.6 Luna (max)OpenAI | 53.59 | SciCode v1.0.1 · Artificial Analysis |
| 43 | GPT-6 Astra (Non-reasoning)OpenAI | 53.47 | SciCode v1.0.1 · Artificial Analysis |
| 44 | Gemini 3.6 Flash (high)Google | 53.36 | SciCode v1.0.1 · Artificial Analysis |
| 45 | Qwen3.8 MaxAlibaba | 53.24 | SciCode v1.0.1 · Artificial Analysis |
| 46 | Grok 4.6 (xhigh)SpaceXAI | 53.01 | SciCode v1.0.1 · Artificial Analysis |
| 47 | Kimi K3 (low)Kimi | 52.66 | SciCode v1.0.1 · Artificial Analysis |
| 48 | GPT-5.6 Terra (high)OpenAI | 52.43 | SciCode v1.0.1 · Artificial Analysis |
| 49 | GPT-5.6 Terra (xhigh)OpenAI | 52.31 | SciCode v1.0.1 · Artificial Analysis |
| 50 | GLM-5.3-FlashZ AI | 51.62 | SciCode v1.0.1 · Artificial Analysis |
| 51 | Claude Opus 5 (Adaptive Reasoning, Medium Effort)Anthropic | 51.5 | SciCode v1.0.1 · Artificial Analysis |
| 52 | GLM-5.2 (max)Z AI | 51.16 | SciCode v1.0.1 · Artificial Analysis |
| 53 | DeepSeek V4 Pro 0813 (Reasoning, Max Effort)DeepSeek | 51.04 | SciCode v1.0.1 · Artificial Analysis |
| 54 | DeepSeek V4 Pro (Reasoning, Max Effort)DeepSeek | 50.81 | SciCode v1.0.1 · Artificial Analysis |
| 55 | MiMo-V2.5-ProXiaomi | 50.58 | SciCode v1.0.1 · Artificial Analysis |
| 56 | Qwen3.8-Flash-NextAlibaba | 50.58 | SciCode v1.0.1 · Artificial Analysis |
| 57 | GPT-5.6 Luna (xhigh)OpenAI | 50.46 | SciCode v1.0.1 · Artificial Analysis |
| 58 | DeepSeek V4 Flash 0731 (Reasoning, Max Effort)DeepSeek | 50.35 | SciCode v1.0.1 · Artificial Analysis |
| 59 | Inkling SmallThinking Machines | 49.65 | SciCode v1.0.1 · Artificial Analysis |
| 60 | DeepSeek V4 Flash Vision (Reasoning, Max Effort)DeepSeek | 49.65 | SciCode v1.0.1 · Artificial Analysis |
| 61 | Qwen3.7 MaxAlibaba | 49.54 | SciCode v1.0.1 · Artificial Analysis |
| 62 | Grok 4.6 (low)SpaceXAI | 49.42 | SciCode v1.0.1 · Artificial Analysis |
| 63 | Claude Opus 5 (Adaptive Reasoning, Low Effort)Anthropic | 49.19 | SciCode v1.0.1 · Artificial Analysis |
| 64 | Agnes 2.5 Pro BetaSapiens AI | 48.84 | SciCode v1.0.1 · Artificial Analysis |
| 65 | Hy3Tencent | 48.61 | SciCode v1.0.1 · Artificial Analysis |
| 66 | Grok 4.3 (high)SpaceXAI | 48.26 | SciCode v1.0.1 · Artificial Analysis |
| 67 | Quasar 438B (max, based on GLM-5.2)Multiverse Computing | 48.15 | SciCode v1.0.1 · Artificial Analysis |
| 68 | Solar Open2 250BUpstage | 48.03 | SciCode v1.0.1 · Artificial Analysis |
| 69 | Kimi K2.7 CodeKimi | 47.8 | SciCode v1.0.1 · Artificial Analysis |
| 70 | MiniMax-M3MiniMax | 47.11 | SciCode v1.0.1 · Artificial Analysis |
| 71 | Inkling (xhigh)Thinking Machines | 46.99 | SciCode v1.0.1 · Artificial Analysis |
| 72 | Qwen3.8 27B (xhigh)Alibaba | 46.64 | SciCode v1.0.1 · Artificial Analysis |
| 73 | Qwen3.7 PlusAlibaba | 46.06 | SciCode v1.0.1 · Artificial Analysis |
| 74 | Apodex 1.1Apodex | 45.49 | SciCode v1.0.1 · Artificial Analysis |
| 75 | Gemma 4 31B (Reasoning)Google | 45.49 | SciCode v1.0.1 · Artificial Analysis |
| 76 | DeepSeek V4 Flash (Reasoning, Max Effort)DeepSeek | 45.25 | SciCode v1.0.1 · Artificial Analysis |
| 77 | Muse Glimmer (high)Meta | 44.91 | SciCode v1.0.1 · Artificial Analysis |
| 78 | Solar Pro 4Upstage | 44.56 | SciCode v1.0.1 · Artificial Analysis |
| 79 | MiMo-V2.5Xiaomi | 43.87 | SciCode v1.0.1 · Artificial Analysis |
| 80 | Step 3.7 FlashStepFun | 43.87 | SciCode v1.0.1 · Artificial Analysis |
| 81 | Gemini 3.1 Flash-LiteGoogle | 43.4 | SciCode v1.0.1 · Artificial Analysis |
| 82 | Nex-N2-ProNex AGI | 43.29 | SciCode v1.0.1 · Artificial Analysis |
| 83 | Agnes 2.5 Pro AlphaSapiens AI | 42.94 | SciCode v1.0.1 · Artificial Analysis |
| 84 | K2 Horizon 375B A23BMBZUAI Institute of Foundation Models | 42.94 | SciCode v1.0.1 · Artificial Analysis |
| 85 | Claude 4.5 Haiku (Reasoning)Anthropic | 42.25 | SciCode v1.0.1 · Artificial Analysis |
| 86 | Motif 3Motif Technologies | 42.25 | SciCode v1.0.1 · Artificial Analysis |
| 87 | K-EXAONE 2.0LG AI Research | 42.01 | SciCode v1.0.1 · Artificial Analysis |
| 88 | Ling 3.0 FlashInclusionAI | 42.01 | SciCode v1.0.1 · Artificial Analysis |
| 89 | Gemini 3.5 Flash-LiteGoogle | 41.32 | SciCode v1.0.1 · Artificial Analysis |
| 90 | A.X-K2SK Telecom | 40.97 | SciCode v1.0.1 · Artificial Analysis |
| 91 | Nemotron 3 Ultra 550B A55B (Reasoning)NVIDIA | 40.28 | SciCode v1.0.1 · Artificial Analysis |
| 92 | Mistral Medium 3.5Mistral | 40.16 | SciCode v1.0.1 · Artificial Analysis |
| 93 | Qwen3.8 27B (low)Alibaba | 40.05 | SciCode v1.0.1 · Artificial Analysis |
| 94 | Qwen3.8 27B (medium)Alibaba | 39 | SciCode v1.0.1 · Artificial Analysis |
| 95 | Command A+Cohere | 38.54 | SciCode v1.0.1 · Artificial Analysis |
| 96 | Granite 4.2 30BIBM | 37.85 | SciCode v1.0.1 · Artificial Analysis |
| 97 | G9v3-39A5BAI9Stars | 36.81 | SciCode v1.0.1 · Artificial Analysis |
| 98 | Mistral Large 3Mistral | 36.57 | SciCode v1.0.1 · Artificial Analysis |
| 99 | LongCat 2.0LongCat | 36.34 | SciCode v1.0.1 · Artificial Analysis |
| 100 | Qwen3.8 27B (Non-reasoning)Alibaba | 36.23 | SciCode v1.0.1 · Artificial Analysis |
| 101 | gpt-oss-120b (high)OpenAI | 34.03 | SciCode v1.0.1 · Artificial Analysis |
| 102 | Nemotron 3.5 LightningNVIDIA | 32.06 | SciCode v1.0.1 · Artificial Analysis |
| 103 | Granite 4.2 8BIBM | 31.48 | SciCode v1.0.1 · Artificial Analysis |
| 104 | Granite 4.2 3BIBM | 25.35 | SciCode v1.0.1 · Artificial Analysis |
| 105 | Ling 3.0 TinyInclusionAI | 24.19 | SciCode v1.0.1 · Artificial Analysis |
| 106 | Celeris-1Celeris | 21.64 | SciCode v1.0.1 · Artificial Analysis |
| 107 | LFM2.5-2.6BLiquid AI | 14.35 | SciCode v1.0.1 · Artificial Analysis |
일치하는 모델이 없습니다. 검색어를 줄여 보세요.
공개된 숫자만 높은 순으로 표시합니다. 순서가 신뢰구간을 고려한 확정 순위는 아닙니다. 과학 연구용 코드 생성의 공개 평가 결과입니다. 일반 저장소 버그 수정이나 앱 구현 점수와 동일하게 해석하지 않습니다.
평가 방법 확인 ↗읽는 법
전체 벤치마크과학 연구용 코드 생성의 공개 평가 결과입니다. 일반 저장소 버그 수정이나 앱 구현 점수와 동일하게 해석하지 않습니다.
과학 연구용 코드 생성의 공개 평가 결과입니다. 일반 저장소 버그 수정이나 앱 구현 점수와 동일하게 해석하지 않습니다.