Model comparison

gemma-4-31B-it-DFlash vs gemma-3-1b-it

At Q4_K_M, gemma-3-1b-it is the smaller download — 806,058,240 bytes against 911,587,296.

From the file· summed bytes, KV per layer

Side by side

gemma-4-31B-it-DFlashgemma-3-1b-it
Parameters1.5B1000M
Architecturedflashgemma3
Layers526
Native context262,14432,768
Mixture of expertsnono
Quantizations published1228
Smallest quantization0.54 GiB0.52 GiB
Q4_K_M0.85 GiB0.75 GiB
Licenceapache-2.0gemma

KV cache by context

the term that decides long-context viability
Contextgemma-4-31B-it-DFlashgemma-3-1b-itRatio
4,0960.05 GiB0.04 GiB1.47×
8,1920.07 GiB0.05 GiB1.33×
16,3840.10 GiB0.08 GiB1.21×
32,7680.16 GiB0.15 GiB1.12×
65,5360.29 GiB0.27 GiB1.06×
131,0720.54 GiB0.52 GiB1.03×