Is IQuest-Coder-V1-40B-Instruct or gemma-4-E4B-it smaller?

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

Which uses less memory at long context?

gemma-4-E4B-it — its KV cache at 32K context is 19.8x smaller, because of how its attention layers cache.

Model comparison

IQuest-Coder-V1-40B-Instruct vs gemma-4-E4B-it

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 19.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	IQuest-Coder-V1-40B-Instruct	gemma-4-E4B-it
Parameters	39.8B	8.0B
Architecture	llama	gemma4
Layers	80	42
Native context	131,072	131,072
Mixture of experts	no	no
Quantizations published	23	36
Smallest quantization	7.91 GiB	3.30 GiB
Q4_K_M	—	4.64 GiB
Licence	other	apache-2.0

KV cache by context

the term that decides long-context viability

Context	IQuest-Coder-V1-40B-Instruct	gemma-4-E4B-it	Ratio
4,096	1.25 GiB	0.12 GiB	10.16×
8,192	2.50 GiB	0.18 GiB	14.07×
16,384	5.00 GiB	0.29 GiB	17.41×
32,768	10.00 GiB	0.51 GiB	19.77×
65,536	20.00 GiB	0.94 GiB	21.20×
131,072	40.00 GiB	1.82 GiB	22.00×