Is GRaPE-2-Pro 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 4.0x smaller, because of how its attention layers cache.

Model comparison

GRaPE-2-Pro 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 4.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	GRaPE-2-Pro	gemma-4-E4B-it
Parameters	27.8B	8.0B
Architecture	qwen35	gemma4
Layers	64	42
Native context	262,144	131,072
Mixture of experts	no	no
Quantizations published	23	36
Smallest quantization	6.66 GiB	3.30 GiB
Q4_K_M	—	4.64 GiB
Licence	apache-2.0	apache-2.0

KV cache by context

the term that decides long-context viability

Context	GRaPE-2-Pro	gemma-4-E4B-it	Ratio
4,096	0.25 GiB	0.12 GiB	2.03×
8,192	0.50 GiB	0.18 GiB	2.81×
16,384	1.00 GiB	0.29 GiB	3.48×
32,768	2.00 GiB	0.51 GiB	3.95×
65,536	4.00 GiB	0.94 GiB	4.24×
131,072	8.00 GiB	1.82 GiB	4.40×