Is gemma-4-12B-it-qat-q4_0-unquantized or Laguna-XS-2.1 smaller?

Both publish quantizations; compare the table for the exact sizes.

Which uses less memory at long context?

Laguna-XS-2.1 — its KV cache at 32K context is 1.8x smaller, because of how its attention layers cache.

Model comparison

gemma-4-12B-it-qat-q4_0-unquantized vs Laguna-XS-2.1

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: Laguna-XS-2.1'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-4-12B-it-qat-q4_0-unquantized	Laguna-XS-2.1
Parameters	12.0B	33.4B
Architecture	gemma4	laguna
Layers	48	40
Native context	262,144	262,144
Mixture of experts	no	yes, 256 experts
Quantizations published	2	31
Smallest quantization	6.50 GiB	8.76 GiB
Q4_K_M	—	18.88 GiB
Licence	apache-2.0	openmdw-1.1

KV cache by context

the term that decides long-context viability

Context	gemma-4-12B-it-qat-q4_0-unquantized	Laguna-XS-2.1	Ratio
4,096	0.72 GiB	0.27 GiB	2.63×
8,192	0.97 GiB	0.43 GiB	2.25×
16,384	1.47 GiB	0.74 GiB	1.98×
32,768	2.47 GiB	1.37 GiB	1.81×
65,536	4.47 GiB	2.62 GiB	1.71×
131,072	8.47 GiB	5.12 GiB	1.65×