Aller au contenu
en Se connecter
8.8 Élevée

CVE-2026-22807 vLLM is an inference and serving engine for large language models (LLMs)

Description

vLLM is an inference and serving engine for large language models (LLMs). Starting in version 0.10.1 and prior to version 0.14.0, vLLM loads Hugging Face `auto_map` dynamic modules during model resolution without gating on `trust_remote_code`, allowing attacker-controlled Python code in a model repo/path to execute at server startup. An attacker who can influence the model repo/path (local directory or remote Hugging Face repo) can achieve arbitrary code execution on the vLLM host during model load. This happens before any request handling and does not require API access. Version 0.14.0 fixes the issue.

Produits et versions touchés

Produit Versions touchées Corrigé en
vllm 0.10 0.10.1 → < 0.14.0 0.14.0
Red Hat AI Inference Server 3.2 1772160625 premières → < 1772160625 1772160625
Red Hat AI Inference Server 3.3 1782352847 premières → < 1782352847 1782352847
Red Hat OpenShift AI 2.25 1783998857 premières → < 1783998857 1783998857
Red Hat OpenShift AI 3.3 1770956034 premières → < 1770956034 1770956034
Red Hat OpenShift AI 3.4 1772204861 premières → < 1772204861 1772204861
Red Hat AI Inference Server Versions non publiées
Red Hat Enterprise Linux AI (RHEL AI) 3 Versions non publiées
Red Hat OpenShift AI (RHOAI) Versions non publiées

Références

Chronologie

  1. Publiée
  2. Dernière modification

Autres CVE de vllm

Les 121 CVE de vllm →
CVE-2026-115918 vLLM is an inference and serving engine for large language models 6.5 Moyenne CVE-2026-115917 vLLM is an inference and serving engine for large language models 3.4 Faible CVE-2026-115916 vLLM is an inference and serving engine for large language models 6.5 Moyenne CVE-2026-115915 vLLM is an inference and serving engine for large language models 8.1 Élevée
Suivre cette CVE