Model comparison
Wan-Dancer-14B vs LTX-2.3
At Q4_K_M, LTX-2.3 is the smaller download — 14,194,441,248 bytes against 26,083,251,744.
From the file· summed bytes, KV per layer
Side by side
| Wan-Dancer-14B | LTX-2.3 | |
|---|---|---|
| Parameters | 17.2B | 9.2B |
| Architecture | wan | qwen35 |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 5 | 92 |
| Smallest quantization | 15.83 GiB | 7.39 GiB |
| Q4_K_M | 24.29 GiB | 13.22 GiB |
| Licence | apache-2.0 | — |
KV cache by context
the term that decides long-context viability
| Context | Wan-Dancer-14B | LTX-2.3 | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |