作者: Jakub Rusinowski · 最后更新: 2026年6月15日
Yes — the Llama 3.1 Family family runs on the NVIDIA GeForce RTX 4060.
| VRAM | 8 GB |
| Memory Bandwidth | 272 GB/s |
| Llama 3.1 8B Instruct | Q4_K_M · 6.5 GB · ~42 tok/s (est.) |
Yes — the Llama 3.1 Family family runs on the NVIDIA GeForce RTX 4060.
Llama 3.1 8B Instruct at Q4_K_M quantization (6.5 GB), estimated ~42 tokens/sec.
← Can I Run It? | Llama 3.1 Family Model Page | NVIDIA GeForce RTX 4060 GPU Page | Check Your Hardware