DeepSeek shipped V4.1-Flash today with MIT open weights — and the interesting number isn't a benchmark score.It keeps 890 bytes of KV cache per token, about a quarter of the previous generation. A 1M-token context now costs ~0.93 GB of cache.For an agent that mostly re-reads its own history, that changes the economics far more than a few benchmark points. #DeepSeek #LLM #OpenSource
Related
FYI: Adobe gains marketing tie-ins with 5 Disney units in multi-year AI alliance: Disney Imagineers already turn sketche...
FYI: Adobe gains marketing tie-ins with 5 Disney units in multi-year AI alliance: Disney Imagineers already turn sketches into concept art with Firefly Foundry models built on Disn...
En un acte de sorpresa extrema, Microsoft ha decidit que és podran desactivar les opcions de IA del navegador Bing.https...
En un acte de sorpresa extrema, Microsoft ha decidit que és podran desactivar les opcions de IA del navegador Bing.https://www.techpowerup.com/354750/microsoft-admits-not-everyone-...
AI Chip Wars 2026: Who Is Really Winning?NVIDIA, TSMC, Broadcom, AMD, and Intel generated a combined $489.7 billion in a...
AI Chip Wars 2026: Who Is Really Winning?NVIDIA, TSMC, Broadcom, AMD, and Intel generated a combined $489.7 billion in annual revenue. But revenue alone doesn't tell the full story...