Is gemma-2-27b-it or gemma-4-E4B-it smaller?

gemma-4-E4B-it is smaller at Q4_K_M: 4.64 GiB against 15.50 GiB.

Which uses less memory at long context?

gemma-4-E4B-it — its KV cache at 32K context is 13.0x smaller, because of how its attention layers cache.

Model comparison

gemma-2-27b-it vs gemma-4-E4B-it

At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 16,645,381,632. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 13.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	gemma-2-27b-it	gemma-4-E4B-it
Parameters	27.2B	8.0B
Architecture	gemma2	gemma4
Layers	46	42
Native context	8,192	131,072
Mixture of experts	no	no
Quantizations published	54	36
Smallest quantization	5.71 GiB	3.30 GiB
Q4_K_M	15.50 GiB	4.64 GiB
Licence	gemma	apache-2.0

KV cache by context

the term that decides long-context viability

Context	gemma-2-27b-it	gemma-4-E4B-it	Ratio
4,096	1.44 GiB	0.12 GiB	11.68×
8,192	2.25 GiB	0.18 GiB	12.64×
16,384	3.68 GiB	0.29 GiB	12.83×
32,768	6.56 GiB	0.51 GiB	12.97×
65,536	12.31 GiB	0.94 GiB	13.05×
131,072	23.81 GiB	1.82 GiB	13.09×