Is gemma-4-E4B-it-ultra-uncensored-heretic or gemma-4-26B-A4B-it smaller?

gemma-4-E4B-it-ultra-uncensored-heretic is smaller at Q4_K_M: 4.97 GiB against 15.64 GiB.

Which uses less memory at long context?

gemma-4-E4B-it-ultra-uncensored-heretic — its KV cache at 32K context is 3.1x smaller, because of how its attention layers cache.

Model comparison

gemma-4-E4B-it-ultra-uncensored-heretic vs gemma-4-26B-A4B-it

At Q4_K_M, gemma-4-E4B-it-ultra-uncensored-heretic is the smaller download — 5,335,286,336 bytes against 16,796,017,248. At long context the gap widens: gemma-4-E4B-it-ultra-uncensored-heretic's KV cache at 32K is 3.1× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	gemma-4-E4B-it-ultra-uncensored-heretic	gemma-4-26B-A4B-it
Parameters	8.0B	26.5B
Architecture	gemma4	gemma4
Layers	42	30
Native context	131,072	262,144
Mixture of experts	no	yes, 128 experts
Quantizations published	24	46
Smallest quantization	4.10 GiB	8.99 GiB
Q4_K_M	4.97 GiB	15.64 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	gemma-4-E4B-it-ultra-uncensored-heretic	gemma-4-26B-A4B-it	Ratio
4,096	0.12 GiB	0.45 GiB	3.65×
8,192	0.18 GiB	0.61 GiB	3.41×
16,384	0.29 GiB	0.92 GiB	3.20×
32,768	0.51 GiB	1.54 GiB	3.05×
65,536	0.94 GiB	2.79 GiB	2.96×
131,072	1.82 GiB	5.29 GiB	2.91×