Side-by-side comparison of Llama 3.1 405B (Meta) and Qwen 2.5 Coder 32B (Alibaba) — benchmarks, pricing, and capabilities.
| Llama 3.1 405B Meta | Qwen 2.5 Coder 32B Alibaba | |
|---|---|---|
| Category | LLMs | Code Models |
| Specifications | ||
| Context Window | 128K | 128K |
| Pricing (per 1M tokens) | ||
| Input Cost | Free | — |
| Output Cost | Free | — |
| Performance | ||
| Overall Score | 88.4 | 88.0 |
| ARC-Challenge | 87.6 | — |
| BigBench Hard | 82.3 | — |
| Chatbot Arena ELO | 1126.0 | — |
| DROP | 82.0 | — |
| GSM8K | 69.7 | — |
| HumanEval | 85.0 | 92.7 |
| LiveCodeBench | — | 42.8 |
| MATH | 70.1 | — |
| MBPP | — | 90.2 |
| MMLU | 84.2 | — |
| SWE-bench Verified | — | 33.5 |
| TruthfulQA | 76.4 | — |
| WinoGrande | 79.4 | — |
| Community | ||
| User Rating | ★ 4.4 | ★ 4.5 |
| Reviews | 420 | 175 |