Model comparison

Nanbeige4.2-3B vs Qwen3-VL-8B-Instruct-abliterated-v1

At Q4_K_M, Nanbeige4.2-3B is the smaller download — 2,574,807,904 bytes against 5,027,785,792.

From the file· summed bytes, KV per layer

Side by side

Nanbeige4.2-3BQwen3-VL-8B-Instruct-abliterated-v1
Parameters4.2B8.8B
Architecturenanbeigeqwen3vl
Layers22
Native context262,144
Mixture of expertsnono
Quantizations published2248
Smallest quantization1.64 GiB1.97 GiB
Q4_K_M2.40 GiB4.68 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextNanbeige4.2-3BQwen3-VL-8B-Instruct-abliterated-v1Ratio
4,0960.34 GiB
8,1920.69 GiB
16,3841.38 GiB
32,7682.75 GiB
65,5365.50 GiB
131,07211.00 GiB
Nanbeige4.2-3B vs Qwen3-VL-8B-Instruct-abliterated-v1 — size, memory and hardware fit — ossmodeldb