FRAMEWIREIndonesiaUpdated Sep 10Live wire
0:00 / 0:00

Ultimate compression KV Cache Compression

What has changed in DeepSeek-V4.1-Flash’s Causal Encoder–Decoder? In a regular Transformer, each layer generates its own K and V based on the hidden states entering this layer. DeepSeek-V4.1-Flash's Causal Encoder–Decoder adjusts this dependency: the decoder's global KV is directly determined by the causal encoder…

Dongxi 东锡 NLPSep 1013
0:00 / 0:00

DeepSeek V4.1 Flash made DOOM in Atomic Chat 🐳

Deepseek_ai dropped a new architecture for the flash model so we tested it by making Doom in an html file and it built a real raycasting shooter with gun mechanics and enemy waves Run AI models locally ->

atomic.chatSep 103
0:00 / 0:00

Three snowy builds. One model.

DeepSeek V4.1 Flash passed our initial market, cabin and Supply Run checks through OpenCode Go. No extra repair round. Three small case studies, with tokens, times and limits in the video. Thanks for watching, fam!

TonkaToyXL | AI Test CabinSep 10
0:00 / 0:00

DeepSeek V4.1 Flash is live on RunBiOS — Day 0.

1M context · 384K output · vision · tool calling · prompt caching at $0.08/M Input $0.40/M · Output $1.60/M. Agentic and coding runs resend context every step, and cached input costs 80% less. Get 50% extra credit on your wallet

UltraSafe AISep 102