Agen Hermes baru saja mengirimkan mode bot di desktop.
Your profiles become a roster of named bots. Each with its own role, model, memory, and avatar. They message each other. Build once, use forever. if you already have Hermes profiles, they show up as bots the moment you update. you
Satu juta token dalam konteksnya digunakan untuk meminta 4 GB VRAM per aliran dalam cache KV saja.
DeepSeek-V4 reduces it to 82 MB: a 98 percent collapse in memory footprint. Here is the three-part architectural co-design that made it possible: 1. Compressed Sparse Attention + FP4
DeepSeek V4 Pro baru saja mencetak skor 87,9 di Terminal Bench 2.1.
That matters because this isn’t another “AI answers questions better” benchmark. It tests whether an AI can actually finish tasks inside a computer terminal. And DeepSeek built V4 Pro around exactly that.
DeepSeek-V4 Pro Max vs Claude Opus 5 vs Grok 4.5 Maks
Testing 2M+ Token Context Retrieval & Complex Code Synthesis: 🔹 DeepSeek-V4 Pro Max: Fastest inference & cost-efficiency for massive codebases (98.4% Needle-In-A-Haystack accuracy). 🔹 Claude 5 Opus: Unrivaled structural
Sudah sampai ujung arsip
Semua deepseek