vLLM before 0.28.0 contains a remote code execution vulnerability in the LlavaOnevision2 processor loader that ignores the trust_remote_code parameter when loading remote processor classes. Attackers can craft a malicious model with arbitrary code in processing_llava_onevision2.py that executes with vLLM process authority even when trust_remote_code is set to False.
https://github.com/vllm-project/vllm/security/advisories/GHSA-3c86-2m5g-59q7
Published: 2026-09-12
Updated: 2026-09-12
Base Score: 7.2
Vector: CVSS2#AV:L/AC:L/Au:N/C:C/I:C/A:C
Severity: High
Base Score: 7.8
Vector: CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Severity: High
Base Score: 8.5
Vector: CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Severity: High