Yes — with 5.5 GB to spare
Qwen3 30B-A3B at Q4_K_M fits your M4 · 32 GB entirely in unified memory at 8K context, at an estimated 17 tokens per second. Past 66K the KV cache pushes it over — quantise the cache to q8_0, or step down a quantisation, to go longer.
Fully in unified memory
8K context
Q4_K_M · 17.1 GB
Apache 2.0
Released Apr 2025
The MoE that made "3B active" a category. Qwen3.6 35B-A3B is its direct replacement.
The VRAM budget
weights 17.1 GB
Weights 17.1 GB
KV cache @ 8K 0.75 GB
Runtime overhead 0.6 GB
Free 5.5 GB of 24.0 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| Q8_0 | 30.2 GB | 31.5 GB | — | ~9.6 | −0.1% ppl | 7.5 GB over |
| Q6_K | 23.3 GB | 24.6 GB | 1K | ~12 | −0.4% ppl | 0.6 GB over |
| Q5_K_M | 20.1 GB | 21.5 GB | 34K | 14 | −0.8% ppl | Long context |
| Q4_K_M | 17.1 GB | 18.5 GB | 66K | 17 | −1.9% ppl | Recommended |
| Q3_K_M | 13.9 GB | 15.2 GB | 101K | 21 | −5.4% ppl | Long context |
| Q2_K | 11.9 GB | 13.2 GB | 122K | 24 | −15% ppl | Long context |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it.
How to run it
$ pip install mlx-lm $ mlx_lm.generate --model mlx-community/Qwen3-30B-A3B-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.
01Download is 17.1 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02macOS caps what the GPU may wire down at about 24.0 GB of 32 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.
03There is room to go to 66K context on this card.