Gemma on Apple M4

Can I run Gemma 3 27B on an Mac Studio M4 Max 128GB?

Yes. Gemma 3 27B at Q4_K_M uses 17.2 GB of the 96.0 GB available on a Mac Studio M4 Max 128GB, and runs at about 21.4 tokens per second. You can push the context to 128K.

Runs comfortably

17.2 GB of 96.0 GB · 18%
096 GB
Weights 15.6 GB
KV cache 1.1 GB
Runtime overhead 0.5 GB

Gemma 3 27B at Q4_K_M leaves 78.8 GB spare on a Mac Studio M4 Max 128GB. There is room to raise the context length or move up a quantisation level.

Generation21.4tok/s
Prompt processing261tok/s
Max context128Ktokens
KV per 1K tokens0GB

Every quantisation of Gemma 3 27B on a Mac Studio M4 Max 128GB

Highlighted row is the highest quality that still fits at 8K context.

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 51.0 GB 52.7 GB Runs comfortably 128K 6.77
INT8 / W8A8 8.50 26.8 GB 28.4 GB Runs comfortably 128K 12.7
Q8_0 (GGUF) 8.50 26.8 GB 28.4 GB Runs comfortably 128K 12.7
FP8 (E4M3) 8.00 25.5 GB 27.1 GB Runs comfortably 128K 13.3
Q6_K 6.56 20.9 GB 22.6 GB Runs comfortably 128K 16.1
Q5_K_M 5.67 18.1 GB 19.7 GB Runs comfortably 128K 18.5
Q5_K_S 5.52 17.6 GB 19.2 GB Runs comfortably 128K 19.0
Q4_K_M 4.85 15.6 GB 17.2 GB Runs comfortably 128K 21.4
AWQ 4-bit 4.25 15.5 GB 17.1 GB Runs comfortably 128K 21.5
GPTQ 4-bit 4.25 15.5 GB 17.1 GB Runs comfortably 128K 21.5
MXFP4 4.25 15.5 GB 17.1 GB Runs comfortably 128K 21.5
Q4_K_S 4.58 14.8 GB 16.4 GB Runs comfortably 128K 22.5
Q4_0 4.55 14.7 GB 16.3 GB Runs comfortably 128K 22.6
IQ4_XS 4.25 13.8 GB 15.4 GB Runs comfortably 128K 24.0
Q3_K_M 3.91 12.7 GB 14.4 GB Runs comfortably 128K 25.8
IQ3_M 3.70 12.1 GB 13.7 GB Runs comfortably 128K 27.1
IQ3_XXS 3.06 10.2 GB 11.8 GB Runs comfortably 128K 31.8
Q2_K 2.63 8.9 GB 10.5 GB Runs comfortably 128K 36.1
IQ2_XXS 2.06 7.1 GB 8.8 GB Runs comfortably 128K 43.9
IQ1_M 1.75 6.2 GB 7.8 GB Runs comfortably 128K 49.7

Also worth checking