Model comparison

Qwen2.5-Omni-3B vs Qwen3-8B

At Q4_K_M, Qwen2.5-Omni-3B is the smaller download — 2,104,931,648 bytes against 5,027,783,488.

From the file· summed bytes, KV per layer

Side by side

Qwen2.5-Omni-3BQwen3-8B
Parameters5.5B8.2B
Architectureqwen2vlqwen3
Layers36
Native context40,960
Mixture of expertsnono
Quantizations published2251
Smallest quantization0.89 GiB2.12 GiB
Q4_K_M1.96 GiB4.68 GiB
Licenceotherapache-2.0

KV cache by context

the term that decides long-context viability
ContextQwen2.5-Omni-3BQwen3-8BRatio
4,0960.56 GiB
8,1921.13 GiB
16,3842.25 GiB
32,7684.50 GiB
65,5369.00 GiB
131,07218.00 GiB