Written by Jakub Rusinowski · Last updated April 22, 2026
Yes — comfortably
Yes, comfortably — Qwen 3.6 35B-A3B at Q8_0 needs about 39.9 GB of the 144 GB usable on Mac Studio (M2 Ultra, 192 GB), leaving ~104.1 GB spare and running at ~80.6 tok/s (estimated), with room for about 262,144 tokens of context.
Confidence: medium · Recommended quantization: Q8_0 · Estimated speed: ~80.6 tok/s
| Usable memory for models | 144 GB |
| Memory bandwidth | 800 GB/s |
| Form factor | Desktop |
| Operating system | macOS |
| Memory upgradeable | No — soldered |
| Quant | Memory needed | Fits 144 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 72.7 GB | ✓ Yes | 256K | ~53.1 tok/s | 70 GB |
| Q8_0 | 39.9 GB | ✓ Yes | 256K | ~80.6 tok/s | 37.2 GB |
| Q6_K | 31.4 GB | ✓ Yes | 256K | ~93 tok/s | 28.7 GB |
| Q5_K_M | 27.5 GB | ✓ Yes | 256K | ~100.1 tok/s | 24.8 GB |
| Q4_K_M | 23.8 GB | ✓ Yes | 256K | ~107.8 tok/s | 21.1 GB |
| Q3_K_M | 17.6 GB | ✓ Yes | 256K | ~124.1 tok/s | 14.9 GB |
| Q2_K | 14.2 GB | ✓ Yes | 256K | ~135.3 tok/s | 11.5 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| Qwen 3.6 35B-A3B | 23.8 GB | ✓ Fits | ~107.8 tok/s |
| Qwen 3.6 27B | 19.3 GB | ✓ Fits | ~23.1 tok/s |
Ollama or LM Studio (Metal) — MLX for the fastest Apple-native throughput
Yes, comfortably — Qwen 3.6 35B-A3B at Q8_0 needs about 39.9 GB of the 144 GB usable on Mac Studio (M2 Ultra, 192 GB), leaving ~104.1 GB spare and running at ~80.6 tok/s (estimated), with room for about 262,144 tokens of context.
Q8_0 — it needs about 39.9 GB of the 144 GB available, downloads as roughly 37.2 GB, and runs at an estimated 80.6 tokens/sec with up to 256K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Ollama or LM Studio (Metal) — MLX for the fastest Apple-native throughput