Is granite-4.0-h-tiny or gemma-4-E4B-it smaller?

granite-4.0-h-tiny is smaller at Q4_K_M: 3.94 GiB against 4.64 GiB.

Which uses less memory at long context?

granite-4.0-h-tiny — its KV cache at 32K context is 2.0x smaller, because of how its attention layers cache.

Model comparison

granite-4.0-h-tiny vs gemma-4-E4B-it

At Q4_K_M, granite-4.0-h-tiny is the smaller download — 4,230,975,936 bytes against 4,977,171,584. At long context the gap widens: granite-4.0-h-tiny's KV cache at 32K is 2.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	granite-4.0-h-tiny	gemma-4-E4B-it
Parameters	6.9B	8.0B
Architecture	granitehybrid	gemma4
Layers	40	42
Native context	131,072	131,072
Mixture of experts	yes, 64 experts	no
Quantizations published	39	36
Smallest quantization	1.81 GiB	3.30 GiB
Q4_K_M	3.94 GiB	4.64 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	granite-4.0-h-tiny	gemma-4-E4B-it	Ratio
4,096	0.03 GiB	0.12 GiB	3.93×
8,192	0.06 GiB	0.18 GiB	2.84×
16,384	0.13 GiB	0.29 GiB	2.30×
32,768	0.25 GiB	0.51 GiB	2.02×
65,536	0.50 GiB	0.94 GiB	1.89×
131,072	1.00 GiB	1.82 GiB	1.82×