Is gemma-4-31B-it-qat-q4_0-unquantized-assistant or Qwen3-0.6B smaller?

gemma-4-31B-it-qat-q4_0-unquantized-assistant is smaller at Q4_K_M: 0.33 GiB against 0.37 GiB.

Which uses less memory at long context?

gemma-4-31B-it-qat-q4_0-unquantized-assistant — its KV cache at 32K context is 6.1x smaller, because of how its attention layers cache.

Model comparison

gemma-4-31B-it-qat-q4_0-unquantized-assistant vs Qwen3-0.6B

At Q4_K_M, gemma-4-31B-it-qat-q4_0-unquantized-assistant is the smaller download — 353,485,920 bytes against 396,705,472. At long context the gap widens: gemma-4-31B-it-qat-q4_0-unquantized-assistant's KV cache at 32K is 6.1× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	gemma-4-31B-it-qat-q4_0-unquantized-assistant	Qwen3-0.6B
Parameters	470M	752M
Architecture	gemma4-assistant	qwen3
Layers	4	28
Native context	262,144	40,960
Mixture of experts	no	no
Quantizations published	13	44
Smallest quantization	0.29 GiB	0.20 GiB
Q4_K_M	0.33 GiB	0.37 GiB
Licence	apache-2.0	—

KV cache by context

the term that decides long-context viability

Context	gemma-4-31B-it-qat-q4_0-unquantized-assistant	Qwen3-0.6B	Ratio
4,096	0.13 GiB	0.44 GiB	3.29×
8,192	0.20 GiB	0.88 GiB	4.48×
16,384	0.32 GiB	1.75 GiB	5.46×
32,768	0.57 GiB	3.50 GiB	6.14×
65,536	1.07 GiB	7.00 GiB	6.54×
131,072	2.07 GiB	14.00 GiB	6.76×