CVEvulnerability

CVE-2026-22807

9.8 CRITICAL Published 2026-01-21 Modified 2026-07-21 Official page ↗
Status
Modified
Published
2026-01-21
Last modified
2026-07-21
Assigner
[email protected]
CVSS 3.1 ([email protected])
8.8 HIGH — CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CVSS 3.1 ([email protected])
9.8 CRITICAL — CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CVSS 3.1 (0b0ca135-0b70-47e7-9f44-1890c2a1c46c)
8.8 HIGH — CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Weaknesses
CWE-94

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.

Affected configurations

• vllm vllm (>= 0.10.1 < 0.14.0)

Related

Weakness · CWE (1)

  • CWE-94 Improper Control of Generation of Code weakness

References