Is grug-35b or gemma-4-26B-A4B-it smaller?

gemma-4-26B-A4B-it is smaller at Q4_K_M: 15.64 GiB against 19.71 GiB.

Which uses less memory at long context?

grug-35b — its KV cache at 32K context is 2.5x smaller, because of how its attention layers cache.

Model comparison

grug-35b vs gemma-4-26B-A4B-it

At Q4_K_M, gemma-4-26B-A4B-it is the smaller download — 16,796,017,248 bytes against 21,166,758,528. At long context the gap widens: grug-35b's KV cache at 32K is 2.5× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	grug-35b	gemma-4-26B-A4B-it
Parameters	35.1B	26.5B
Architecture	qwen35moe	gemma4
Layers	40	30
Native context	262,144	262,144
Mixture of experts	yes, 256 experts	yes, 128 experts
Quantizations published	30	46
Smallest quantization	9.11 GiB	8.99 GiB
Q4_K_M	19.71 GiB	15.64 GiB
Licence	mit	apache-2.0

KV cache by context

the term that decides long-context viability

Context	grug-35b	gemma-4-26B-A4B-it	Ratio
4,096	0.08 GiB	0.45 GiB	5.75×
8,192	0.16 GiB	0.61 GiB	3.87×
16,384	0.31 GiB	0.92 GiB	2.94×
32,768	0.63 GiB	1.54 GiB	2.47×
65,536	1.25 GiB	2.79 GiB	2.23×
131,072	2.50 GiB	5.29 GiB	2.12×