Model comparison

Skywork-R1V3-38B vs Qwen3.6-35B-A3B

At Q4_K_M, Skywork-R1V3-38B is the smaller download — 19,848,566,240 bytes against 21,166,757,728.

From the file· summed bytes, KV per layer

Side by side

Skywork-R1V3-38BQwen3.6-35B-A3B
Parameters38.4B36.0B
Architectureqwen2qwen35moe
Layers40
Native context262,144
Mixture of expertsnoyes, 256 experts
Quantizations published3061
Smallest quantization8.41 GiB8.77 GiB
Q4_K_M18.49 GiB19.71 GiB
Licencemitapache-2.0

KV cache by context

the term that decides long-context viability
ContextSkywork-R1V3-38BQwen3.6-35B-A3BRatio
4,0960.08 GiB
8,1920.16 GiB
16,3840.31 GiB
32,7680.63 GiB
65,5361.25 GiB
131,0722.50 GiB