Can XYZ-Aquila mini 35B-A3B run on MacBook Pro M3 Pro 18GB?
NO — Won’t fit
Q4_K_M · 4K context · formula estimate
What this means
The checked Q4_K_M build needs about 25.7GB, above this Mac's conservative 15GB working budget.
−10.7GB short
above the working budget
25.7GB
needed at Q4_K_M
Q4_K_M
quant selected
0GB
working headroom
needs 25.7 GBworking budget 15 GB · short 10.7 GB
Where the memory goes
| Component | Detail | GB |
|---|---|---|
| Model weights | Q4_K_M GGUF (21.4 GB) + mmap overhead | 22.5 |
| KV cache | 4K context window | 2.5 |
| Runtime | macOS inference app + compute buffers | 0.8 |
| Total needed | at Q4_K_M, 4K context | 25.7 |
| Working budget | 18 GB unified memory − conservative macOS reserve | 15 |
| Headroom | memory shortfall | −10.7 |
Pick your quant
| Quant | Download | Memory | Estimated speed | Verdict |
|---|---|---|---|---|
| Q2_K | 12.6 GB | 16.5 GB | — | ✕ Won't fit |
| Q3_K_M | 16.2 GB | 20.3 GB | — | ✕ Won't fit |
| IQ4_XS | 18.8 GB | 23 GB | — | ✕ Won't fit |
| Q4_0 | 19.9 GB | 24.1 GB | — | ✕ Won't fit |
| Q4_K_S | 20.6 GB | 24.9 GB | — | ✕ Won't fit |
| Q4_K_M ★ | 21.4 GB | 25.7 GB | — | ✕ Won't fit |
| Q5_K_M | 25 GB | 29.5 GB | — | ✕ Won't fit |
| Q6_K | 30.1 GB | 34.9 GB | — | ✕ Won't fit |
| Q8_0 | 36.9 GB | 42 GB | — | ✕ Won't fit |
Other models on this Mac
XYZ-Aquila mini 35B-A3B on other MacBook Pro M3 Pro configurations
FAQ
Can the MacBook Pro M3 Pro · 18GB run XYZ-Aquila mini 35B-A3B?
Not with the quants currently tracked. The selected Q4_K_M build needs about 25.7GB, above the 15GB working budget.
Which XYZ-Aquila mini 35B-A3B quant should I use on this Mac?
None of the tracked quants fit this configuration safely. Choose a smaller model or a Mac with more unified memory.
Which app should I use for XYZ-Aquila mini 35B-A3B on this Mac?
Start with LM Studio. This page gives the complete point-and-click walkthrough.
Are these speeds measured on a MacBook Pro M3 Pro?
No. The range is a formula estimate based on memory bandwidth, model size, active parameters, cooling, and a 4K context. The app, backend, thermals, and prompt can change real performance.