what/llm/can/i/run
gpus/apple/m3-max-64gb

M3 Max 64GB

Released 2023-10-30 · Street $3,499
vram
64 GB
unified mem
bandwidth
400
GB/s
models fit
18
top · Q8_0
power
78 W
low

Spec sheet · top pick

Hardware spec
M3 Max 64GB
Apple · Apple Silicon · M3
vram
64 GB unified
bandwidth
400 GB/s
tdp
78 W
released
2023-10-30
msrp
$3,499
street
~$3,499
★ Sweet spot for this card
Qwen2.5 Coder 32B
Q8_0 · GGUF · 34.1 GB
bench48.8 · conf A
vram34.1 / 64 GB
speed12 tok/s
Best benchmark per GB of VRAM in your tier. Q8_0wastes <3% quality vs FP16, leaves headroom for 32k context.
Full model inspect →

Top 10 models on this GPU

#modelquantvramtok/sbench ↓fit
01Qwen2.5 Coder 32BQ8_034.1 GB1248.8Great
02gemma 3 27bFP1657.2 GB750.6Ok
03Qwen3 32BQ8_034.1 GB1240.0Great
04Gemma 2 9BFP1619.4 GB2135.0Great
05Gemma 2 27BFP1657.2 GB746.2Ok
06deepseek r1 distill llama 70bQ6_K55.6 GB751.7Ok
07Gemma 3 12BFP1625.7 GB1631.3Great
08deepseek r1 distill qwen 32bQ8_034.1 GB1233.3Great
09Gemma 3 4BFP168.9 GB4515.1Great
10Mixtral 8x22b v0.1Q2_K56 GB720.0Ok
Showing 10 of 18 models that fit · see full ranked list →

Upgrade options

★ Best value
+H100 PCIe
80 GB · ~$28,000 · +$24,501 over M3 Max 64GB
Same model set · higher headroom for context / batch · top-pick bench +1.8
See H100 PCIe
Bigger tier
+H100 SXM5
80 GB · ~$40,000 · +$36,501 over M3 Max 64GB
Same model set · higher headroom for context / batch · top-pick bench +1.8
See H100 SXM5
Premium tier
+H200 SXM5
141 GB · ~$45,000 · +$41,501 over M3 Max 64GB
Unlocks +1 models not in current rank · top-pick bench +1.8
See H200 SXM5

Also compare

A6000 Pro
48 GB · NVIDIA · 20% more
RTX 6000 Ada
48 GB · NVIDIA · 100% more
H100 PCIe
80 GB · NVIDIA · 700% more
H100 SXM5
80 GB · NVIDIA · 1043% more
Not running a M3 Max 64GB?
Tell us what you actually have — we'll re-rank everything for your machine.
Change my machine