Is Phi-4-reasoning or gemma-4-E4B-it smaller?

gemma-4-E4B-it is smaller at Q4_K_M: 4.64 GiB against 8.43 GiB.

Which uses less memory at long context?

gemma-4-E4B-it — its KV cache at 32K context is 12.4x smaller, because of how its attention layers cache.

Model comparison

Phi-4-reasoning vs gemma-4-E4B-it

At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 9,053,116,096. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 12.4× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Phi-4-reasoning	gemma-4-E4B-it
Parameters	14.7B	8.0B
Architecture	phi3	gemma4
Layers	40	42
Native context	32,768	131,072
Mixture of experts	no	no
Quantizations published	46	36
Smallest quantization	3.33 GiB	3.30 GiB
Q4_K_M	8.43 GiB	4.64 GiB
Licence	mit	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Phi-4-reasoning	gemma-4-E4B-it	Ratio
4,096	0.78 GiB	0.12 GiB	6.35×
8,192	1.56 GiB	0.18 GiB	8.79×
16,384	3.13 GiB	0.29 GiB	10.88×
32,768	6.25 GiB	0.51 GiB	12.36×
65,536	12.50 GiB	0.94 GiB	13.25×
131,072	25.00 GiB	1.82 GiB	13.75×