Is gemma-4-12B-it-qat-q4_0-unquantized or gemma-2-9b-it smaller?

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

Which uses less memory at long context?

gemma-4-12B-it-qat-q4_0-unquantized — its KV cache at 32K context is 2.4x smaller, because of how its attention layers cache.

Model comparison

gemma-4-12B-it-qat-q4_0-unquantized vs gemma-2-9b-it

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: gemma-4-12B-it-qat-q4_0-unquantized's KV cache at 32K is 2.4× 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	gemma-2-9b-it
Parameters	12.0B	9.2B
Architecture	gemma4	gemma2
Layers	48	42
Native context	262,144	8,192
Mixture of experts	no	no
Quantizations published	2	46
Smallest quantization	6.50 GiB	2.86 GiB
Q4_K_M	—	5.37 GiB
Licence	apache-2.0	gemma

KV cache by context

the term that decides long-context viability

Context	gemma-4-12B-it-qat-q4_0-unquantized	gemma-2-9b-it	Ratio
4,096	0.72 GiB	1.31 GiB	1.83×
8,192	0.97 GiB	2.05 GiB	2.12×
16,384	1.47 GiB	3.36 GiB	2.29×
32,768	2.47 GiB	5.99 GiB	2.43×
65,536	4.47 GiB	11.24 GiB	2.51×
131,072	8.47 GiB	21.74 GiB	2.57×