Model comparison

InternVL3_5-30B-A3B vs Qwen3.5-9B

At Q4_K_M, Qwen3.5-9B is the smaller download — 5,627,044,256 bytes against 18,632,180,352.

From the file· summed bytes, KV per layer

Side by side

InternVL3_5-30B-A3BQwen3.5-9B
Parameters30.8B9.7B
Architectureqwen3moeqwen35
Layers32
Native context262,144
Mixture of expertsnono
Quantizations published3034
Smallest quantization7.05 GiB2.97 GiB
Q4_K_M17.35 GiB5.24 GiB
Licenceapache-2.0

KV cache by context

the term that decides long-context viability
ContextInternVL3_5-30B-A3BQwen3.5-9BRatio
4,0960.13 GiB
8,1920.25 GiB
16,3840.50 GiB
32,7681.00 GiB
65,5362.00 GiB
131,0724.00 GiB