p PatchFirst

CVE-2026-22807

CVE-2026-22807 · vLLM affected by RCE via auto_map dynamic module loading during model initialization

This page contains facts from public vulnerability sources. Check the original advisory before updating.

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.

CVSS 8.8 · High · EPSS 1.0% · Fix available

Published
Updated

Stack: python

Affected products and versions

  • vllm

    Versions
    • >= 0.10.1, < 0.14.0: Affected

    No fix published

  • vllm PyPI

    Versions
    • 0.10.1
    • 0.10.1.1
    • 0.10.2
    • 0.11.0
    • 0.11.1
    • 0.11.2
    • 0.12.0
    • 0.13.0
    Affected ranges
    • ECOSYSTEM: ≥ 0.10.1, < 0.14.0

    Update to 0.14.0

Original sources

Check your server →