Autor: Jakub Rusinowski · Ostatnia aktualizacja: 15 czerwca 2026
Yes — the Llama 3.1 Family family runs on the NVIDIA GeForce RTX 4070 Ti Super.
| VRAM | 16 GB |
| Memory Bandwidth | 672 GB/s |
| Llama 3.1 8B Instruct | Q4_K_M · 6.5 GB · ~103 tok/s (est.) |
Yes — the Llama 3.1 Family family runs on the NVIDIA GeForce RTX 4070 Ti Super.
Llama 3.1 8B Instruct at Q4_K_M quantization (6.5 GB), estimated ~103 tokens/sec.
← Can I Run It? | Llama 3.1 Family Model Page | NVIDIA GeForce RTX 4070 Ti Super GPU Page | Check Your Hardware