Model comparison

MiniMax-H3 vs Sulphur-2-base

At Q4_K_M, Sulphur-2-base is the smaller download — 5,629,108,608 bytes against 14,576,977,960.

From the file· summed bytes, KV per layer

Side by side

MiniMax-H3Sulphur-2-base
Parameters33.1B9.2B
Architectureqwen3vlltxv
Layers
Native context
Mixture of expertsnono
Quantizations published2213
Smallest quantization10.60 GiB5.07 GiB
Q4_K_M13.58 GiB5.24 GiB
Licenceother

KV cache by context

the term that decides long-context viability
ContextMiniMax-H3Sulphur-2-baseRatio
4,096
8,192
16,384
32,768
65,536
131,072