BaselineWatch
Desk
Your fleet and your work Fleetyour releases and the hardware they run onTriagewhat needs a decisionWatchlistthe products and platforms you runSaved searchesthe lists you come back to
This instance Sourceswhat is read, how often, under what termsDigestthe daily or weekly summary, and alerts
How it is drawn Appearancethe look and the colour schemeBrandingthe mark this instance wears
Owner sign-in

CVE-2026-41523

7.5 high SSVC exploitation: none Modified
NVD · CVE.org

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLLM runs in Python optimized mode (python -O or PYTHONOPTIMIZE=1). This vulnerability is fixed in 0.22.0.

Fix

Patch, vendor advisory and mitigation references on NVD 1

Every line above is quoted from NVD, CISA, Microsoft, the vendor's own advisory or DISA; nothing is generated.

Change history 1

  1. First seenCVSS 7.5 · vllmNVD modified 2026-10-25 23:10

Affected (CPE criteria) 1

  • cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*

References 20