Llama 3.3 70B Instruct vs Llama 3.1 70B Instruct
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Llama 3.3 70B Instruct | Llama 3.1 70B Instruct |
|---|---|---|
| Provider |  Meta | Meta |
| score moyen | 46.9 | 37.8 |
| Prix d'entrée | $0.10 | $0.40 |
| Prix de sortie | $0.32 | $0.40 |
| Fenêtre de contexte | 131K tokens (~66 books) | 131K tokens (~66 books) |
| Sorti le | 2024-12-06 | 2024-07-23 |
| Code source ouvert | Open Source | Open Source |
Scores de benchmark
8 benchmarks · Llama 3.3 70B Instruct: 6, Llama 3.1 70B Instruct: 2
| Benchmark | Catégorie | Llama 3.3 70B Instruct | Llama 3.1 70B Instruct |
|---|---|---|---|
| Aider — Code Editing | coding | 59.4 | 58.6 |
| BBH (HuggingFace) | general | 56.6 | 55.9 |
| Chatbot Arena Elo — Overall | arena | 1318.0 | 1292.8 |
| GPQA | knowledge | 10.5 | 14.2 |
| IFEval | language | 90.0 | 86.7 |
| MATH Level 5 | math | 48.3 | 38.1 |
| MMLU-PRO | knowledge | 48.1 | 47.9 |
| MUSR | reasoning | 15.6 | 17.7 |