Qwen2.5 32B vs Mistral Small 3: Benchmark Comparison
Detailed comparison of Qwen2.5 32B and Mistral Small 3 covering benchmarks, pricing, context window, and compliance.
Key Specifications
| Specification | Qwen2.5 32B | Mistral Small 3 |
|---|---|---|
| Vendor | alibaba | mistral |
| Version | 2.5-32b | small-3 |
| Release Date | 2024-09-19 | 2025-01-29 |
| Context Window | 131072 tokens | 32768 tokens |
| Input Modalities | text | text |
| Output Modalities | text | text |
| License | Qwen License | Apache 2.0 |
| SOC2 | ✗ | ✓ |
| HIPAA | ✗ | ✗ |
| GDPR | ✗ | ✓ |
| ISO 27001 | ✗ | ✓ |
Benchmark Results
| Benchmark | Qwen2.5 32B | Mistral Small 3 | Winner |
|---|---|---|---|
| ARC | 93.9 | 91.8 | Qwen2.5 32B |
| BBH | 76.8 | 71.3 | Qwen2.5 32B |
| GPQA | 39.4 | 30.1 | Qwen2.5 32B |
| GSM8K | 76.3 | 79.6 | Mistral Small 3 |
| HUMANEVAL | 68 | 74.4 | Mistral Small 3 |
| IFEVAL | 73.5 | 76.9 | Mistral Small 3 |
| MATH | 39.5 | 39.2 | Qwen2.5 32B |
| MMLU | 78.5 | 76.9 | Qwen2.5 32B |
| MUSR | 54.9 | 46 | Qwen2.5 32B |
| WINOGRANDE | 81.9 | 78.8 | Qwen2.5 32B |
Pricing Comparison
| Tier (per Mtok) | Qwen2.5 32B | Mistral Small 3 |
|---|---|---|
| Input | $0.35 | $0.2 |
| Output | $0.45 | $0.6 |
| Cache Read | $0 | $0 |
| Cache Write | $0 | $0 |
Qwen2.5 32B 대 Mistral Small 3
모델 개요
Qwen2.5 32B and Mistral Small 3 are both notable options in the AI model market. This page compares their benchmarks, pricing, and compliance.
핵심 사양
| 공급업체 | 출시일 | 컨텍스트 창 | 라이선스 |
|---|---|---|---|
| Alibaba / Mistral | 2024-09-19 / 2025-01-29 | 131K / 32K | Qwen License / Apache 2.0 |
벤치마크 성능
| 벤치마크 | Qwen2.5 32B | Mistral Small 3 | 승자 |
|---|---|---|---|
| ARC | 93.9 | 91.8 | A |
| BBH (BIG-Bench Hard) | 76.8 | 71.3 | A |
| GPQA | 39.4 | 30.1 | A |
| GSM8K (Grade School Math 8K) | 76.3 | 79.6 | B |
| HumanEval | 68.0 | 74.4 | B |
| IFEval | 73.5 | 76.9 | B |
| MATH | 39.5 | 39.2 | Tie |
| MMLU (Massive Multitask Language Understanding) | 78.5 | 76.9 | A |
| MUSR | 54.9 | 46.0 | A |
| WinoGrande | 81.9 | 78.8 | A |
가격 비교
| 입력 | 출력 | 캐시 읽기 | 캐시 쓰기 |
|---|---|---|---|
| — / — | — / — | — / — | — / — |
백만 토큰당 — A / B
강점 & 약점
Qwen2.5 32B
- ✅ 可靠的通用模型。
- ⚠️ 闭源专有模型，不支持自托管。
Mistral Small 3
- ✅ 可靠的通用模型。
- ⚠️ 闭源专有模型，不支持自托管。
편집자 의견
Qwen2.5 32B and Mistral Small 3 each have their strengths. Choose based on workload (code, long context, vision), referencing the tables above.
FAQ
Which model is better for coding tasks?
Refer to the HumanEval benchmark table; the model with a higher score is better suited for coding tasks.
Which model is cheaper?
Refer to the pricing comparison table above; the model with lower input/output prices is more cost-effective.
Which has a longer context window?
Refer to the key specifications table; the model with a larger context window is better for long documents.
참고문헌
Editor's Take
See Editor's Take section.