Model comparison

Qwen3-Coder-30B-A3B-Instruct vs translategemma-27b-it

At Q4_K_M, translategemma-27b-it is the smaller download — 16,546,704,032 bytes against 18,556,689,568.

From the file· summed bytes, KV per layer

Side by side

Qwen3-Coder-30B-A3B-Instructtranslategemma-27b-it
Parameters30.5B28.8B
Architectureqwen3moegemma3
Layers48
Native context262,144
Mixture of expertsyes, 128 expertsno
Quantizations published4641
Smallest quantization7.46 GiB5.83 GiB
Q4_K_M17.28 GiB15.41 GiB
Licenceapache-2.0gemma

KV cache by context

the term that decides long-context viability
ContextQwen3-Coder-30B-A3B-Instructtranslategemma-27b-itRatio
4,0960.38 GiB
8,1920.75 GiB
16,3841.50 GiB
32,7683.00 GiB
65,5366.00 GiB
131,07212.00 GiB