Llama on Apple M3
Can I run Llama 4 Scout 109B-A17B on an Mac Studio M3 Ultra 256GB?
Yes. Llama 4 Scout 109B-A17B at Q4_K_M uses 63.7 GB of the 192.0 GB available on a Mac Studio M3 Ultra 256GB, and runs at about 38.5 tokens per second. You can push the context to 692K.
Runs comfortably
63.7 GB of 192.0 GB · 33%0192 GB
Weights
61.7 GB
KV cache
1.5 GB
Runtime overhead
0.5 GB
Llama 4 Scout 109B-A17B at Q4_K_M leaves 128.3 GB spare on a Mac Studio M3 Ultra 256GB. There is room to raise the context length or move up a quantisation level.
Generation38.5tok/s
Prompt processing704tok/s
Max context692Ktokens
KV per 1K tokens0GB
Every quantisation of Llama 4 Scout 109B-A17B on a Mac Studio M3 Ultra 256GB
Highlighted row is the highest quality that still fits at 8K context.
| Quantisation | bpw | Weights | Total | Verdict | Max ctx | tok/s |
|---|---|---|---|---|---|---|
| FP16 / BF16 | 16.00 | 203.0 GB | 205.1 GB | Won't fit | — | 6.81 |
| INT8 / W8A8 | 8.50 | 107.4 GB | 109.4 GB | Runs comfortably | 448K | 23.0 |
| Q8_0 (GGUF) | 8.50 | 107.4 GB | 109.4 GB | Runs comfortably | 448K | 23.0 |
| FP8 (E4M3) | 8.00 | 101.5 GB | 103.6 GB | Runs comfortably | 480K | 24.2 |
| Q6_K | 6.56 | 83.2 GB | 85.3 GB | Runs comfortably | 577K | 29.2 |
| Q5_K_M | 5.67 | 71.9 GB | 74.0 GB | Runs comfortably | 637K | 33.4 |
| Q5_K_S | 5.52 | 70.0 GB | 72.1 GB | Runs comfortably | 648K | 34.3 |
| Q4_K_M | 4.85 | 61.7 GB | 63.7 GB | Runs comfortably | 692K | 38.5 |
| Q4_K_S | 4.58 | 58.3 GB | 60.4 GB | Runs comfortably | 710K | 40.5 |
| Q4_0 | 4.55 | 58.0 GB | 60.0 GB | Runs comfortably | 712K | 40.7 |
| AWQ 4-bit | 4.25 | 56.8 GB | 58.8 GB | Runs comfortably | 718K | 41.5 |
| GPTQ 4-bit | 4.25 | 56.8 GB | 58.8 GB | Runs comfortably | 718K | 41.5 |
| MXFP4 | 4.25 | 56.8 GB | 58.8 GB | Runs comfortably | 718K | 41.5 |
| IQ4_XS | 4.25 | 54.2 GB | 56.3 GB | Runs comfortably | 732K | 43.2 |
| Q3_K_M | 3.91 | 50.0 GB | 52.0 GB | Runs comfortably | 754K | 46.5 |
| IQ3_M | 3.70 | 47.4 GB | 49.4 GB | Runs comfortably | 768K | 48.8 |
| IQ3_XXS | 3.06 | 39.4 GB | 41.5 GB | Runs comfortably | 811K | 57.3 |
| Q2_K | 2.63 | 34.1 GB | 36.1 GB | Runs comfortably | 839K | 64.9 |
| IQ2_XXS | 2.06 | 27.0 GB | 29.0 GB | Runs comfortably | 877K | 78.8 |
| IQ1_M | 1.75 | 23.1 GB | 25.1 GB | Runs comfortably | 898K | 89.2 |