Is gemma-4-26B-A4B-it or EXAONE-4.5-33B smaller?

gemma-4-26B-A4B-it is smaller at Q4_K_M: 15.64 GiB against 18.67 GiB.

Which uses less memory at long context?

gemma-4-26B-A4B-it — its KV cache at 32K context is 1.8x smaller, because of how its attention layers cache.

Model comparison

gemma-4-26B-A4B-it vs EXAONE-4.5-33B

At Q4_K_M, gemma-4-26B-A4B-it is the smaller download — 16,796,017,248 bytes against 20,047,839,424. At long context the gap widens: gemma-4-26B-A4B-it's KV cache at 32K is 1.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	gemma-4-26B-A4B-it	EXAONE-4.5-33B
Parameters	26.5B	34.4B
Architecture	gemma4	exaone4
Layers	30	64
Native context	262,144	262,144
Mixture of experts	yes, 128 experts	no
Quantizations published	46	31
Smallest quantization	8.99 GiB	11.57 GiB
Q4_K_M	15.64 GiB	18.67 GiB
Licence	apache-2.0	other

KV cache by context

the term that decides long-context viability

Context	gemma-4-26B-A4B-it	EXAONE-4.5-33B	Ratio
4,096	0.45 GiB	1.00 GiB	2.23×
8,192	0.61 GiB	1.34 GiB	2.22×
16,384	0.92 GiB	1.84 GiB	2.01×
32,768	1.54 GiB	2.84 GiB	1.84×
65,536	2.79 GiB	4.84 GiB	1.73×
131,072	5.29 GiB	8.84 GiB	1.67×