DeepSeek V4 Flash 284B (13B active)
on NVIDIA RTX 4060 Ti 16GB · llama.cpp · 131,072 ctx
Sep 7, 2026
Summary
User reports a 98 GB model running on 16 GB VRAM with CPU offload, improving from 2 to 7 t/s with recent llama.cpp updates.
on NVIDIA RTX 4060 Ti 16GB · llama.cpp · 131,072 ctx