SaibaRadar

CVE-2026-114053 — Vulnerability intelligence

vllm — vllm

Published:

CVSS
9.2
CISA KEV
Not listed; not proof of no exploitation

Original source text (not a translation)

vLLM from 0.7.3 before 0.31.0 forwards the request-supplied mm_processor_kwargs.code_revision to Transformers AutoProcessor.from_pretrained, letting clients select which revision of a model's dynamic processor code is imported when trust_remote_code is enabled. An attacker controlling a revision in the served model repository can send a multimodal request to the unauthenticated /tokenize endpoint to execute arbitrary Python in the vLLM API process.

Sources