VulnerabilityModified

CVE-2026-22807

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…

CRITICAL 9.8EPSS 0.76%

Does this matter?

High impact if exploited, but EPSS currently rates exploitation as unlikely (0.76%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit.

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.

CVSS 3.1: 9.8 CRITICALCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
EPSS: 0.76% probability · 53th percentile
CISA KEV: Not listed
Weakness: CWE-94
Affected: vllm/vllm
Source: security-advisories@github.com

References

Source: NVD record, EPSS from FIRST.org, KEV from CISA. Refreshed daily.