Back to registry
CUDA192GB VRAM

arnav080/deepseek-v4-flash-hikari

DeepSeek V4-Flash MoE (284B) production profile tuned for 2x RTX 6000 GPUs

Configuration Specifications

Base Modelhuggingface:deepseek-ai/DeepSeek-V4-Flash
Enginellama.cpp
QuantizationQ4_K_M
Platformcuda
VRAM Required192GB minimum

Telemetry Benchmarks

No runs recorded yet

Be the first to benchmark this recipe! Run the CLI command from your terminal:

bloc run arnav080/deepseek-v4-flash-hikari
deepseek-v4-flash-hikari.yaml
Loading workspace editor...