Is GLM-4-32B-0414 or gemma-4-E4B-it smaller?

gemma-4-E4B-it is smaller at Q4_K_M: 4.64 GiB against 18.33 GiB.

Which uses less memory at long context?

gemma-4-E4B-it — its KV cache at 32K context is 3.8x smaller, because of how its attention layers cache.

Model comparison

GLM-4-32B-0414 vs gemma-4-E4B-it

At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 19,680,022,112. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 3.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	GLM-4-32B-0414	gemma-4-E4B-it
Parameters	32.6B	8.0B
Architecture	glm4	gemma4
Layers	61	42
Native context	32,768	131,072
Mixture of experts	no	no
Quantizations published	49	36
Smallest quantization	7.17 GiB	3.30 GiB
Q4_K_M	18.33 GiB	4.64 GiB
Licence	mit	apache-2.0

KV cache by context

the term that decides long-context viability

Context	GLM-4-32B-0414	gemma-4-E4B-it	Ratio
4,096	0.24 GiB	0.12 GiB	1.94×
8,192	0.48 GiB	0.18 GiB	2.68×
16,384	0.95 GiB	0.29 GiB	3.32×
32,768	1.91 GiB	0.51 GiB	3.77×
65,536	3.81 GiB	0.94 GiB	4.04×
131,072	7.63 GiB	1.82 GiB	4.19×