Model comparison

Chroma vs Krea-2-Turbo

At Q4_K_M, Chroma is the smaller download — 5,566,533,792 bytes against 7,216,993,376.

From the file· summed bytes, KV per layer

Side by side

ChromaKrea-2-Turbo
Parameters8.9B12.8B
Architecturefluxqwen_image
Layers
Native context
Mixture of expertsnono
Quantizations published27333
Smallest quantization3.99 GiB2.78 GiB
Q4_K_M5.18 GiB6.72 GiB
Licenceapache-2.0other

KV cache by context

the term that decides long-context viability
ContextChromaKrea-2-TurboRatio
4,096
8,192
16,384
32,768
65,536
131,072