Model comparison

gemma-4-12B-it vs MR_MARTIN_31b_V1.0_Gemma-4-base-ftp

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

gemma-4-12B-itMR_MARTIN_31b_V1.0_Gemma-4-base-ftp
Parameters12.0B30.7B
Architecturegemma4gemma4
Layers48
Native context262,144
Mixture of expertsnono
Quantizations published5923
Smallest quantization3.92 GiB6.66 GiB
Q4_K_M6.63 GiB
Licenceapache-2.0other

KV cache by context

the term that decides long-context viability
Contextgemma-4-12B-itMR_MARTIN_31b_V1.0_Gemma-4-base-ftpRatio
4,0960.72 GiB
8,1920.97 GiB
16,3841.47 GiB
32,7682.47 GiB
65,5364.47 GiB
131,0728.47 GiB
gemma-4-12B-it vs MR_MARTIN_31b_V1.0_Gemma-4-base-ftp — size, memory and hardware fit — ossmodeldb