Is ThinkingCap-Qwen3.6-27B-heretic or gemma-4-12B-it smaller?

gemma-4-12B-it is smaller at Q4_K_M: 6.63 GiB against 15.41 GiB.

Which uses less memory at long context?

ThinkingCap-Qwen3.6-27B-heretic — its KV cache at 32K context is 1.2x smaller, because of how its attention layers cache.

Model comparison

ThinkingCap-Qwen3.6-27B-heretic vs gemma-4-12B-it

At Q4_K_M, gemma-4-12B-it is the smaller download — 7,121,861,440 bytes against 16,547,399,328. At long context the gap widens: ThinkingCap-Qwen3.6-27B-heretic's KV cache at 32K is 1.2× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	ThinkingCap-Qwen3.6-27B-heretic	gemma-4-12B-it
Parameters	27.4B	12.0B
Architecture	qwen35	gemma4
Layers	64	48
Native context	262,144	262,144
Mixture of experts	no	no
Quantizations published	17	59
Smallest quantization	8.72 GiB	3.92 GiB
Q4_K_M	15.41 GiB	6.63 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	ThinkingCap-Qwen3.6-27B-heretic	gemma-4-12B-it	Ratio
4,096	0.25 GiB	0.72 GiB	2.87×
8,192	0.50 GiB	0.97 GiB	1.94×
16,384	1.00 GiB	1.47 GiB	1.47×
32,768	2.00 GiB	2.47 GiB	1.23×
65,536	4.00 GiB	4.47 GiB	1.12×
131,072	8.00 GiB	8.47 GiB	1.06×