How fast is Qwen3 30B-A3B on an Mac?

Around 74 tokens per second. Token generation on Apple Silicon is limited by memory bandwidth, and this machine has 300 GB/s. Because this is a mixture-of-experts model only 3.3B of its 30.5B parameters are read per token, which is why it is faster than its size suggests.

How much memory does Qwen3 30B-A3B need?

17.4 GiB for the weights at Q4_K_M, plus 0.8 GiB for an 8K-token KV cache and about 1.7 GiB of runtime overhead — 19.8 GiB in total.

Home › Qwen3 30B-A3B › M3 Max 36GB

Can an M3 Max 36GB Mac run Qwen3 30B-A3B?

M3 Max · 36 GB unified memory · 300 GB/s · MacBook Pro 14" (M3 Max), MacBook Pro 16" (M3 Max)

Yes — Qwen3 30B-A3B runs on an M3 Max 36GB Mac.

At Q4_K_M it needs about 19.8 GiB, leaving 10.8 GiB spare out of the ~30.6 GiB macOS will let you use. Expect roughly 74 tokens/sec — faster than you can read.

Weights

17.4 GiB

30.5B params @ Q4_K_M

KV cache

0.8 GiB

8K context, FP16

Overhead

1.7 GiB

runtime + activations

Total needed

19.8 GiB

of ~30.6 GiB usable

Est. speed

74 tok/s

300 GB/s bandwidth

Max context

64K tokens

at Q4_K_M

Where the memory goes

Three things occupy memory when a model is loaded: the weights themselves, the KV cache that holds the conversation, and the runtime's own working buffers. Only the first is fixed — the KV cache grows with every token in your context.

Quantisation	Weights	Total	Tok/s	Fits?
FP16	56.8 GiB	61.2 GiB	23	No
Q8_0	30.2 GiB	33.2 GiB	43	No
Q6_K	23.4 GiB	26.2 GiB	55	Yes, but tight
Q5_K_M	20.2 GiB	22.8 GiB	64	Yes, comfortably
Q4_K_M	17.4 GiB	19.8 GiB	74	Yes, comfortably
Q3_K_M	13.8 GiB	16.1 GiB	93	Yes, comfortably
Q2_K	10.7 GiB	12.7 GiB	121	Yes, comfortably

The default. Best quality-per-gigabyte for most people.

How long a context fits

The KV cache is often what breaks a setup that looked fine at load time. Quantising it to 8-bit roughly halves its footprint with little measurable quality cost.

Context	KV cache	Total	FP16 KV	Q8 KV
2K	0.2 GiB	19.0 GiB	Yes, comfortably	Yes, comfortably
4K	0.4 GiB	19.3 GiB	Yes, comfortably	Yes, comfortably
8K	0.8 GiB	19.8 GiB	Yes, comfortably	Yes, comfortably
16K	1.5 GiB	20.9 GiB	Yes, comfortably	Yes, comfortably
32K	3.0 GiB	23.0 GiB	Yes, comfortably	Yes, comfortably
128K	12.0 GiB	35.6 GiB	No	Yes, but tight

How to run it

Ollama is the shortest path. It picks a quantisation automatically — usually Q4_K_M, which is what the numbers above assume.

brew install ollama
ollama serve &
ollama run qwen3:30b-a3b

For MLX — Apple's own array framework, typically a little faster than llama.cpp on Apple Silicon:

pip install mlx-lm
mlx_lm.generate --model mlx-community/Qwen3-30B-A3B-4bit --prompt "Hello"

Macs that run Qwen3 30B-A3B comfortably

The smallest configuration that handles it is the M1 Max 32GB (MacBook Pro 14" (M1 Max)).

Mac	Memory	Bandwidth	Tok/s
M1 Max 32GB	32 GB	400 GB/s	99
M2 Max 32GB	32 GB	400 GB/s	99
M4 32GB	32 GB	120 GB/s	30
M3 Max 36GB	36 GB	300 GB/s	74
M4 Max 36GB	36 GB	410 GB/s	101
M4 Pro 48GB	48 GB	273 GB/s	68
M4 Max 48GB	48 GB	546 GB/s	135
M1 Max 64GB	64 GB	400 GB/s	99
M2 Ultra 64GB	64 GB	800 GB/s	198
M4 Pro 64GB	64 GB	273 GB/s	68
M4 Max 64GB	64 GB	546 GB/s	135
M2 Max 96GB	96 GB	400 GB/s	99

Common questions

Can an M3 Max 36GB Mac run Qwen3 30B-A3B?: Yes. At Q4_K_M it needs about 19.8 GiB of the roughly 30.6 GiB available, generating around 74 tokens per second.
How fast is Qwen3 30B-A3B on an Mac?: Around 74 tokens per second. Token generation on Apple Silicon is limited by memory bandwidth, and this machine has 300 GB/s. Because this is a mixture-of-experts model only 3.3B of its 30.5B parameters are read per token, which is why it is faster than its size suggests.
How much memory does Qwen3 30B-A3B need?: 17.4 GiB for the weights at Q4_K_M, plus 0.8 GiB for an 8K-token KV cache and about 1.7 GiB of runtime overhead — 19.8 GiB in total.