THREATOPS
THREAT OPS › Threat News › [NVD] CVE-2026-48746 (CRITICAL 9.1) — 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

[NVD] CVE-2026-48746 (CRITICAL 9.1) — 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

lownvdPublished 2026-06-22

CVE-2026-48746 CVSS: 9.1 CRITICAL Published: 2026-06-22T23:16:30.490

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 -

Indicators of compromise

Original source: https://nvd.nist.gov/vuln/detail/CVE-2026-48746