Model comparison

FLUX.2-klein-9B vs FLUX.2-klein-4B

At Q4_K_M, FLUX.2-klein-4B is the smaller download — 2,497,280,416 bytes against 5,027,783,648.

From the file· summed bytes, KV per layer

Side by side

FLUX.2-klein-9BFLUX.2-klein-4B
Parameters9.1B3.9B
Architecturefluxflux
Layers
Native context
Mixture of expertsnono
Quantizations published2921
Smallest quantization3.71 GiB1.70 GiB
Q4_K_M4.68 GiB2.33 GiB
Licenceotherapache-2.0

KV cache by context

the term that decides long-context viability
ContextFLUX.2-klein-9BFLUX.2-klein-4BRatio
4,096
8,192
16,384
32,768
65,536
131,072