Is Mira-v1.12-Ties-27B 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 6.1x smaller, because of how its attention layers cache.

Model comparison

Mira-v1.12-Ties-27B 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 6.1× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Mira-v1.12-Ties-27B	gemma-4-E4B-it
Parameters	27.4B	8.0B
Architecture	gemma3	gemma4
Layers	62	42
Native context	131,072	131,072
Mixture of experts	no	no
Quantizations published	23	36
Smallest quantization	5.83 GiB	3.30 GiB
Q4_K_M	—	4.64 GiB
Licence	—	apache-2.0

KV cache by context

the term that decides long-context viability

Context	Mira-v1.12-Ties-27B	gemma-4-E4B-it	Ratio
4,096	0.92 GiB	0.12 GiB	7.49×
8,192	1.23 GiB	0.18 GiB	6.95×
16,384	1.86 GiB	0.29 GiB	6.48×
32,768	3.11 GiB	0.51 GiB	6.15×
65,536	5.61 GiB	0.94 GiB	5.95×
131,072	10.61 GiB	1.82 GiB	5.83×