Which uses less memory at long context?

gemma-4-12B-it-qat-q4_0-unquantized — its KV cache at 32K context is 1.7x smaller, because of how its attention layers cache.

Model comparison

Qwen3-42B-A3B-2507-Thinking-Abliterated-uncensored-TOTAL-RECALL-v2-Medium-MASTER-CODER vs gemma-4-12B-it-qat-q4_0-unquantized

Q: Is Qwen3-42B-A3B-2507-Thinking-Abliterated-uncensored-TOTAL-RECALL-v2-Medium-MASTER-CODER or gemma-4-12B-it-qat-q4_0-unquantized smaller?

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

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 1.7× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Qwen3-42B-A3B-2507-Thinking-Abliterated-uncensored-TOTAL-RECALL-v2-Medium-MASTER-CODER	gemma-4-12B-it-qat-q4_0-unquantized
Parameters	42.4B	12.0B
Architecture	qwen3moe	gemma4
Layers	67	48
Native context	262,144	262,144
Mixture of experts	yes, 128 experts	no
Quantizations published	34	2
Smallest quantization	8.21 GiB	6.50 GiB
Q4_K_M	23.94 GiB	—
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Qwen3-42B-A3B-2507-Thinking-Abliterated-uncensored-TOTAL-RECALL-v2-Medium-MASTER-CODER	gemma-4-12B-it-qat-q4_0-unquantized	Ratio
4,096	0.52 GiB	0.72 GiB	1.37×
8,192	1.05 GiB	0.97 GiB	1.08×
16,384	2.09 GiB	1.47 GiB	1.43×
32,768	4.19 GiB	2.47 GiB	1.70×
65,536	8.38 GiB	4.47 GiB	1.87×
131,072	16.75 GiB	8.47 GiB	1.98×