ai model report

Is nvidia/DeepSeek-V4-Flash-NVFP4 safe to load?

CLEAN risk. 3 of 162 tracked dependencies carry advisories. Driven by No version-matched advisories and no malicious packages..

View on HuggingFace

// posture

fig. 01 — posture on eight axes, 100 is clean. Weakest: code execution at 70.

// the finding

0.0xyz score out of 10CLEAN

INFOconfig.json was authored against transformers 4.57.1, which has known advisories. Consumers normally install a current version — this only bites if you pin for reproducibility.

1 more finding on this model.

Author: nvidia
Task: text-generation
Loader: Model Optimizer
License: mit
Downloads: 106,853
Remote code: not required

// dependencies

13 direct, 8 with findings, 155 reachable

Package	Evidence	Advisories	Worst
transformers	direct, runtime	23	none pinned
torch	direct, runtime	13	none pinned
tqdm	direct, install	2	hidden
fast-hadamard-transform	direct, install	0	hidden
nvidia-modelopt	direct, install	0	hidden
safetensors	direct, install	0	hidden
tilelang	direct, install	0	hidden
annotated-doc	direct, install	0	hidden

Create a free accountto see all 162 dependencies, which versions are pinned, and what to change.

Risk computed 2026-09-22 at 18:59 UTC. Dependencies come from the model's requirements, its declared loader and the code it ships; advisories from NVD, GHSA and OSV.