Model comparison

Wan2.2-S2V-14B vs Wan2.2-I2V-A14B

At Q4_K_M, Wan2.2-I2V-A14B is the smaller download — 9,651,728,896 bytes against 13,861,697,120.

From the file· summed bytes, KV per layer

Side by side

Wan2.2-S2V-14BWan2.2-I2V-A14B
Parameters16.3B14.3B
Architecturewanwan
Layers
Native context
Mixture of expertsnono
Quantizations published3030
Smallest quantization8.86 GiB4.94 GiB
Q4_K_M12.91 GiB8.99 GiB
Licenceapache-2.0apache-2.0

KV cache by context

the term that decides long-context viability
ContextWan2.2-S2V-14BWan2.2-I2V-A14BRatio
4,096
8,192
16,384
32,768
65,536
131,072