Model comparison

gemma-4-26B-A4B-it vs gemma-3-27b-it-qat-q4_0-unquantized

At Q4_K_M, gemma-3-27b-it-qat-q4_0-unquantized is the smaller download — 16,546,688,736 bytes against 16,796,017,248.

From the file· summed bytes, KV per layer

Side by side

gemma-4-26B-A4B-itgemma-3-27b-it-qat-q4_0-unquantized
Parameters26.5B27.4B
Architecturegemma4gemma3
Layers30
Native context262,144
Mixture of expertsyes, 128 expertsno
Quantizations published4645
Smallest quantization8.99 GiB6.06 GiB
Q4_K_M15.64 GiB15.41 GiB
Licenceapache-2.0gemma

KV cache by context

the term that decides long-context viability
Contextgemma-4-26B-A4B-itgemma-3-27b-it-qat-q4_0-unquantizedRatio
4,0960.45 GiB
8,1920.61 GiB
16,3840.92 GiB
32,7681.54 GiB
65,5362.79 GiB
131,0725.29 GiB
gemma-4-26B-A4B-it vs gemma-3-27b-it-qat-q4_0-unquantized — size, memory and hardware fit — ossmodeldb