DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression

 (zartbot.github.io)

37 points | by mfiguiere 1 hour ago

2 comments

  • arikrahman 23 minutes ago
    I am very impressed with the KV Cache Compression work as well as the prefix cacheing making queries converge on practically free.
  • smy20011 10 minutes ago
    Should we flag this since It's AI generated?