Model comparison

lingbot-world-v2-14b-causal-fast vs LTX2.3-10Eros

At Q4_K_M, lingbot-world-v2-14b-causal-fast is the smaller download — 12,054,188,608 bytes against 14,296,160,672.

From the file· summed bytes, KV per layer

Side by side

lingbot-world-v2-14b-causal-fastLTX2.3-10Eros
Parameters18.5B21.0B
Architecturewanltxv
Layers
Native context
Mixture of expertsnono
Quantizations published536
Smallest quantization11.23 GiB9.63 GiB
Q4_K_M11.23 GiB13.31 GiB
Licencecc-by-nc-sa-4.0

KV cache by context

the term that decides long-context viability
Contextlingbot-world-v2-14b-causal-fastLTX2.3-10ErosRatio
4,096
8,192
16,384
32,768
65,536
131,072