Is Yi-34B-200K-Llamafied 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 14.8x smaller, because of how its attention layers cache.

Model comparison

Yi-34B-200K-Llamafied 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 14.8× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	Yi-34B-200K-Llamafied	gemma-4-E4B-it
Parameters	34.4B	8.0B
Architecture	llama	gemma4
Layers	60	42
Native context	200,000	131,072
Mixture of experts	no	no
Quantizations published	21	36
Smallest quantization	6.98 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	Yi-34B-200K-Llamafied	gemma-4-E4B-it	Ratio
4,096	0.94 GiB	0.12 GiB	7.62×
8,192	1.88 GiB	0.18 GiB	10.55×
16,384	3.75 GiB	0.29 GiB	13.06×
32,768	7.50 GiB	0.51 GiB	14.83×
65,536	15.00 GiB	0.94 GiB	15.90×
131,072	30.00 GiB	1.82 GiB	16.50×