Model comparison

Nethena-20B vs gemma-4-E4B-it

At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 12,042,208,800. At long context the gap widens: gemma-4-E4B-it's KV cache at 32K is 76.6× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Nethena-20Bgemma-4-E4B-it
Parameters20.0B8.0B
Architecturellamagemma4
Layers6242
Native context4,096131,072
Mixture of expertsnono
Quantizations published1236
Smallest quantization7.74 GiB3.30 GiB
Q4_K_M11.22 GiB4.64 GiB
Licencecc-by-nc-4.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextNethena-20Bgemma-4-E4B-itRatio
4,0964.84 GiB0.12 GiB39.37×
8,1929.69 GiB0.18 GiB54.51×
16,38419.38 GiB0.29 GiB67.48×
32,76838.75 GiB0.51 GiB76.60×
65,53677.50 GiB0.94 GiB82.15×
131,072155.00 GiB1.82 GiB85.24×