Benchmark Performance
| Benchmark | Score | Unit | Evaluated At | Notes | Source |
|---|
| MMLU | 78.2 | % | 2024-09-12 | 5-shot | view |
| HUMANEVAL | 78 | pass@1 | 2024-09-12 | — | view |
| GSM8K | 86.6 | % | 2024-09-12 | 0-shot CoT | view |
| MATH | 44.2 | % | 2024-09-12 | 0-shot CoT | view |
| BBH | 77.2 | % | 2024-09-12 | 3-shot CoT | view |
| GPQA | 40.8 | % | 2024-09-12 | 0-shot | view |
| IFEVAL | 73.1 | % | 2024-09-12 | prompt_strict | view |
| ARC | 93.7 | % | 2024-09-12 | challenge | view |
| MUSR | 54 | % | 2024-09-12 | 0-shot | view |
| WINOGRANDE | 78.3 | % | 2024-09-12 | 0-shot | view |
Pricing
| Tier | Price | Currency |
|---|
| Input | $3 / Mtok | USD |
| Output | $12 / Mtok | USD |
| Cache Read | $0 / Mtok | USD |
| Cache Write | $0 / Mtok | USD |
Source:
https://openai.com/api/pricing/
· as of 2024-09-12
Compliance
- Data Residency: US
- SOC2: ✓
- HIPAA: ✓
- GDPR: ✓
- ISO 27001: ✓
o1 mini
모델 개요
OpenAI o1-mini 经济型推理模型, 针对编程与数学优化, 比 o1 便宜 80%, 适合 STEM 任务。
핵심 사양
| 공급업체 | 버전 | 출시일 | 컨텍스트 창 | 입력 모달리티 | 출력 모달리티 | 라이선스 |
|---|
| Openai | o1-mini | 2024-09-12 | 128K | text | text | Proprietary |
벤치마크 성능
| 벤치마크 | 점수 | 단위 | 비고 |
|---|
| MMLU (Massive Multitask Language Understanding) | 78.2 | % | 5-shot |
| HumanEval | 78.0 | pass@1 | — |
| GSM8K (Grade School Math 8K) | 86.6 | % | 0-shot CoT |
| MATH | 44.2 | % | 0-shot CoT |
| BBH (BIG-Bench Hard) | 77.2 | % | 3-shot CoT |
| GPQA | 40.8 | % | 0-shot |
| IFEval | 73.1 | % | prompt_strict |
| ARC | 93.7 | % | challenge |
| MUSR | 54.0 | % | 0-shot |
| WinoGrande | 78.3 | % | 0-shot |
가격
백만 토큰당
강점
- GSM8K 86.6, robust math reasoning.
약점
사용 사례
참고문헌