Model comparison
FLUX.2-klein-base-9B vs Krea-2-Turbo
At Q4_K_M, FLUX.2-klein-base-9B is the smaller download — 5,909,829,920 bytes against 7,216,993,376.
From the file· summed bytes, KV per layer
Side by side
| FLUX.2-klein-base-9B | Krea-2-Turbo | |
|---|---|---|
| Parameters | 9.1B | 12.8B |
| Architecture | flux | qwen_image |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 15 | 33 |
| Smallest quantization | 3.71 GiB | 2.78 GiB |
| Q4_K_M | 5.50 GiB | 6.72 GiB |
| Licence | cc-by-nc-4.0 | other |
KV cache by context
the term that decides long-context viability
| Context | FLUX.2-klein-base-9B | Krea-2-Turbo | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |