Model comparison
LTX2.3-10Eros vs Wan2.2-Animate-14B
At Q4_K_M, Wan2.2-Animate-14B is the smaller download — 11,496,331,072 bytes against 14,296,160,672.
From the file· summed bytes, KV per layer
Side by side
| LTX2.3-10Eros | Wan2.2-Animate-14B | |
|---|---|---|
| Parameters | 21.0B | 17.3B |
| Architecture | ltxv | wan |
| Layers | — | — |
| Native context | — | — |
| Mixture of experts | no | no |
| Quantizations published | 36 | 17 |
| Smallest quantization | 9.63 GiB | 6.01 GiB |
| Q4_K_M | 13.31 GiB | 10.71 GiB |
| Licence | — | apache-2.0 |
KV cache by context
the term that decides long-context viability
| Context | LTX2.3-10Eros | Wan2.2-Animate-14B | Ratio |
|---|---|---|---|
| 4,096 | — | — | — |
| 8,192 | — | — | — |
| 16,384 | — | — | — |
| 32,768 | — | — | — |
| 65,536 | — | — | — |
| 131,072 | — | — | — |