Claude Opus 4.1
제공 Anthropic · 출시일 2025-08-05
41.3
평균 점수
$15.00/1M
입력 가격
$75.00/1M
출력 가격
200K tokens (~100 books)
컨텍스트 윈도우
multimodal
유형
Tested on 14 benchmarks with 41.3% average. Top scores: Lech Mazur Writing (85.4%), SWE-Bench verified (73.3%), GPQA diamond (69.7%).
벤치마크 점수
| 벤치마크 | 카테고리 | 점수 | Bar |
|---|---|---|---|
| Lech Mazur Writing | knowledge | 85.4 | |
| SWE-Bench verified | coding | 73.3 | |
| GPQA diamond | knowledge | 69.7 | |
| OTIS Mock AIME 2024-2025 | math | 68.9 | |
| SimpleBench | reasoning | 52.0 | |
| DeepResearch Bench | knowledge | 49.7 | |
| WeirdML | coding | 42.8 | |
| Cybench | coding | 42.0 | |
| Terminal Bench | coding | 38.0 | |
| SimpleQA Verified | knowledge | 34.8 | |
| FrontierMath-2025-02-28-Private | math | 7.2 | |
| HLE | knowledge | 7.1 | |
| FrontierMath-Tier-4-2025-07-01-Private | math | 4.2 | |
| VPCT | knowledge | 2.5 |
유사 모델
Alibaba Qwen
41.3
Google DeepMind
41.3
Mistral AI
41.2
OpenAI
41.5