[CVE-2026-48746] vLLM is an inference and serving engine for large language models (LLMs). From 0.3.0 until 0.22.0, a…

Descripción técnica

vLLM is an inference and serving engine for large language models (LLMs). From 0.3.0 until 0.22.0, a vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware. It allows to use the API without providing the configured VLLM_API_KEY or --api-key. This vulnerability is fixed in 0.22.0.

Puntuación CVSS

Score: 9.1/10 — Severidad: CRITICAL — Estado NIST: Modified

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:H

Productos afectados

Vllm — Vllm

Debilidades (CWE)

CWE-444, CWE-501

Fuente oficial

Publicado en NIST NVD.