ai model report

Is moonshotai/Kimi-K2-Thinking safe to load?

LOW risk. 4 of 105 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 25.

// the finding

0.0xyz score out of 10LOW

MEDIUMconfig.json declares `auto_map` — from_pretrained(trust_remote_code=True) executes Python shipped in this repo.

2 more findings on this model.

Author: moonshotai
Task: text-generation
Loader: transformers
License: other
Downloads: 53,978
Remote code: ships Python that runs on load (trust_remote_code)

// dependencies

15 direct, 10 with findings, 97 reachable

Package	Evidence	Advisories	Worst
transformers	direct, runtime	23	none pinned
torch	direct, runtime	13	none pinned
numpy	direct, install	8	hidden
flash-attn	direct, install	1	hidden
huggingface-hub	direct, install	0	hidden
safetensors	direct, install	0	hidden
tiktoken	direct, install	0	hidden
tokenizers	direct, install	0	hidden

Create a free accountto see all 105 dependencies, which versions are pinned, and what to change.

// what changed

MEDIUMrisk level raised (unknown to low)2026-09-09

Risk computed 2026-09-22 at 18:58 UTC. Dependencies come from the model's requirements, its declared loader and the code it ships; advisories from NVD, GHSA and OSV.