Model comparison

DeepSeek-V4-Flash vs openPangu-2.0-Flash

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-FlashopenPangu-2.0-Flash
Parameters291B100B
Architecturedeepseek4openpangu-v2
Layers4346
Native context1,048,576524,288
Mixture of expertsyes, 256 expertsyes, 256 experts
Quantizations published125
Smallest quantization76.87 GiB40.73 GiB
Q4_K_M51.48 GiB
Licencemitother

KV cache by context

the term that decides long-context viability
ContextDeepSeek-V4-FlashopenPangu-2.0-FlashRatio
4,0960.06 GiB
8,1920.06 GiB
16,3840.06 GiB
32,7680.06 GiB
65,5360.06 GiB
131,0720.06 GiB