DeepSeek V3.2 Base Experimental - new Chinese model is an absolute banger ๐Ÿค–
DeepSeek V3.2 Base Experimental - new Chinese model is an absolute banger ๐Ÿค–
Very cheap long context for LLMs:
~3.5x cheaper prefill
~10x cheaper decode
at 128k context at inference with the same quality
0
0 comments
Piotr Macai
1
DeepSeek V3.2 Base Experimental - new Chinese model is an absolute banger ๐Ÿค–
powered by
AinsiderAI
skool.com/ainsider-6149
Build your own community
Bring people together around your passion and get paid.
Powered by