Written by Jakub Rusinowski · Last updated July 10, 2026
Yes — the Llama 3.2 Family family runs on the Apple M1 Ultra.
| VRAM | 128 GB unified memory |
| Memory Bandwidth | 800 GB/s |
| Llama 3.2 90B Vision Instruct | Q4_K_M · 54 GB · ~15 tok/s (est.) |
| Llama 3.2 11B Vision Instruct | Q4_K_M · 7.8 GB · ~103 tok/s (est.) |
| Llama 3.2 3B Instruct | Q4_K_M · 2.2 GB · ~364 tok/s (est.) |
| Llama 3.2 1B Instruct | Q4_K_M · 0.8 GB · ~400 tok/s (est.) |
Yes — the Llama 3.2 Family family runs on the Apple M1 Ultra.
Llama 3.2 90B Vision Instruct at Q4_K_M quantization (54 GB), estimated ~15 tokens/sec.
← Can I Run It? | Llama 3.2 Family Model Page | Apple M1 Ultra GPU Page | Check Your Hardware