Model comparison

Gemma-4-31B-Isometry-Fabled-Persona vs gpt-oss-20b

At Q4_K_M, gpt-oss-20b is the smaller download — 11,624,759,488 bytes against 18,687,065,824. At long context the gap widens: gpt-oss-20b's KV cache at 32K is 8.0× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

Gemma-4-31B-Isometry-Fabled-Personagpt-oss-20b
Parameters31.3B21.5B
Architecturegemma4gpt-oss
Layers6024
Native context262,144131,072
Mixture of expertsnoyes, 32 experts
Quantizations published3415
Smallest quantization6.67 GiB10.68 GiB
Q4_K_M17.40 GiB10.83 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextGemma-4-31B-Isometry-Fabled-Personagpt-oss-20bRatio
4,0961.80 GiB0.11 GiB16.14×
8,1922.42 GiB0.21 GiB11.81×
16,3843.67 GiB0.39 GiB9.35×
32,7686.17 GiB0.77 GiB8.04×
65,53611.17 GiB1.52 GiB7.36×
131,07221.17 GiB3.02 GiB7.02×