Model comparison

DeepSeek-V4-Flash vs MiMo-V2.5

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

DeepSeek-V4-FlashMiMo-V2.5
Parameters291B311B
Architecturedeepseek4mimo2
Layers4348
Native context1,048,5761,048,576
Mixture of expertsyes, 256 expertsyes, 256 experts
Quantizations published1244
Smallest quantization76.87 GiB60.09 GiB
Q4_K_M175.80 GiB
Licencemitmit

KV cache by context

the term that decides long-context viability
ContextDeepSeek-V4-FlashMiMo-V2.5Ratio
4,0960.06 GiB
8,1920.06 GiB
16,3840.06 GiB
32,7680.06 GiB
65,5360.06 GiB
131,0720.06 GiB