Model comparison

gemma-4-E4B-it-The-DECKARD-Claude-Opus-Expresso-Universe-HERETIC-UNCENSORED-Thinking vs gemma-4-E4B-it

At Q4_K_M, gemma-4-E4B-it is the smaller download — 4,977,171,584 bytes against 5,335,286,688.

From the file· summed bytes, KV per layer

Side by side

gemma-4-E4B-it-The-DECKARD-Claude-Opus-Expresso-Universe-HERETIC-UNCENSORED-Thinkinggemma-4-E4B-it
Parameters8.0B8.0B
Architecturegemma4gemma4
Layers4242
Native context131,072131,072
Mixture of expertsnono
Quantizations published2736
Smallest quantization4.10 GiB3.30 GiB
Q4_K_M4.97 GiB4.64 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
Contextgemma-4-E4B-it-The-DECKARD-Claude-Opus-Expresso-Universe-HERETIC-UNCENSORED-Thinkinggemma-4-E4B-itRatio
4,0960.12 GiB0.12 GiB1.00×
8,1920.18 GiB0.18 GiB1.00×
16,3840.29 GiB0.29 GiB1.00×
32,7680.51 GiB0.51 GiB1.00×
65,5360.94 GiB0.94 GiB1.00×
131,0721.82 GiB1.82 GiB1.00×