How fast is Llama 3.1 405B on an Mac?

Around 0 tokens per second. Token generation on Apple Silicon is limited by memory bandwidth, and this machine has 68 GB/s.

How much memory does Llama 3.1 405B need?

231.0 GiB for the weights at Q4_K_M, plus 3.9 GiB for an 8K-token KV cache and about 12.4 GiB of runtime overhead — 247.3 GiB in total.

Home › Llama 3.1 405B › M1 16GB

Can an M1 16GB Mac run Llama 3.1 405B?

M1 · 16 GB unified memory · 68 GB/s · MacBook Air (M1), Mac mini (M1)

No — not at Q4_K_M.

Llama 3.1 405B needs about 247.3 GiB but an M1 16GB Mac has only ~13.0 GiB available for a model — you are 234.3 GiB short. macOS will swap to SSD and generation will crawl.

Weights

231.0 GiB

405B params @ Q4_K_M

KV cache

3.9 GiB

8K context, FP16

Overhead

12.4 GiB

runtime + activations

Total needed

247.3 GiB

of ~13.0 GiB usable

Est. speed

0 tok/s

68 GB/s bandwidth

Max context

n/a

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	Fits?
FP16	754.4 GiB	796.8 GiB	No
Q8_0	400.8 GiB	425.5 GiB	No
Q6_K	311.2 GiB	331.5 GiB	No
Q5_K_M	268.7 GiB	286.9 GiB	No
Q4_K_M	231.0 GiB	247.3 GiB	No
Q3_K_M	183.9 GiB	197.8 GiB	No
Q2_K	141.4 GiB	153.3 GiB	No

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	1.0 GiB	244.1 GiB	No	No
4K	2.0 GiB	245.2 GiB	No	No
8K	3.9 GiB	247.3 GiB	No	No
16K	7.9 GiB	251.6 GiB	No	No
32K	15.8 GiB	260.0 GiB	No	No
128K	63.0 GiB	310.9 GiB	No	No

What to run instead on an M1 16GB Mac

These are the largest models that fit comfortably on this machine at Q4_K_M with an 8K context.

Model	Size	Total	Tok/s
Qwen3 14B	14.8B	10.9 GiB	6	Yes, comfortably
Gemma 2 9B	9.24B	9.0 GiB	10	Yes, comfortably
Qwen3 8B	8.2B	6.8 GiB	11	Yes, comfortably
Llama 3.1 8B	8.03B	6.6 GiB	11	Yes, comfortably
DeepSeek-R1-Distill-Qwen-7B	7.62B	5.8 GiB	12	Yes, comfortably
Qwen2.5 7B	7.62B	5.8 GiB	12	Yes, comfortably

Macs that run Llama 3.1 405B comfortably

The smallest configuration that handles it is the M3 Ultra 512GB (Mac Studio (M3 Ultra)).

Mac	Memory	Bandwidth	Tok/s
M3 Ultra 512GB	512 GB	819 GB/s	3

Common questions

Can an M1 16GB Mac run Llama 3.1 405B?: No. At Q4_K_M it needs about 247.3 GiB, and an M1 16GB Mac has roughly 13.0 GiB available for a model after macOS takes its share.
How fast is Llama 3.1 405B on an Mac?: Around 0 tokens per second. Token generation on Apple Silicon is limited by memory bandwidth, and this machine has 68 GB/s.
How much memory does Llama 3.1 405B need?: 231.0 GiB for the weights at Q4_K_M, plus 3.9 GiB for an 8K-token KV cache and about 12.4 GiB of runtime overhead — 247.3 GiB in total.