Is Qwen3.6-35B-A3B or gemma-4-31B-it-uncensored-heretic smaller?

gemma-4-31B-it-uncensored-heretic is smaller at Q4_K_M: 17.40 GiB against 19.71 GiB.

Which uses less memory at long context?

Qwen3.6-35B-A3B — its KV cache at 32K context is 9.9x smaller, because of how its attention layers cache.

Model comparison

Qwen3.6-35B-A3B vs gemma-4-31B-it-uncensored-heretic

At Q4_K_M, gemma-4-31B-it-uncensored-heretic is the smaller download — 18,687,058,368 bytes against 21,166,757,728. At long context the gap widens: Qwen3.6-35B-A3B's KV cache at 32K is 9.9× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3.6-35B-A3B	gemma-4-31B-it-uncensored-heretic
Parameters	36.0B	31.3B
Architecture	qwen35moe	gemma4
Layers	40	60
Native context	262,144	262,144
Mixture of experts	yes, 256 experts	no
Quantizations published	61	22
Smallest quantization	8.77 GiB	11.10 GiB
Q4_K_M	19.71 GiB	17.40 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Qwen3.6-35B-A3B	gemma-4-31B-it-uncensored-heretic	Ratio
4,096	0.08 GiB	1.80 GiB	22.98×
8,192	0.16 GiB	2.42 GiB	15.49×
16,384	0.31 GiB	3.67 GiB	11.75×
32,768	0.63 GiB	6.17 GiB	9.87×
65,536	1.25 GiB	11.17 GiB	8.94×
131,072	2.50 GiB	21.17 GiB	8.47×