Model comparison

DeepSeek-V4-Flash vs Nemotron-3-Super-49B-v1

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-FlashNemotron-3-Super-49B-v1
Parameters291B121B
Architecturedeepseek4nemotron_h_moe
Layers43
Native context1,048,576
Mixture of expertsyes, 256 expertsno
Quantizations published127
Smallest quantization76.87 GiB49.83 GiB
Q4_K_M80.14 GiB
Licencemitother

KV cache by context

the term that decides long-context viability
ContextDeepSeek-V4-FlashNemotron-3-Super-49B-v1Ratio
4,0960.06 GiB
8,1920.06 GiB
16,3840.06 GiB
32,7680.06 GiB
65,5360.06 GiB
131,0720.06 GiB