CVE-2026-100648

[CVE catalog]

Severity

Medium · CVSS 6.9CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Published

Sep 13, 2026, 10:00:00 PM UTC

Updated

Sep 26, 2026, 2:16:47 PM UTC

Source

[NVD] · [CVE record]

EPSS

0.35% · Oct 25, 2026

A flaw was found in vllm. This vulnerability allows a remote, unauthenticated attacker to cause a Denial of Service (DoS) by bypassing file size restrictions during multimodal chat audio processing. By submitting requests with oversized audio files, an attacker can force the server to allocate excessive memory and Central Processing Unit (CPU) resources during decoding. vllm: vllm: Denial of Service via oversized audio decoding This vulnerability is rated as Moderate severity because, although an unauthenticated remote caller can induce excessive resource exhaustion resulting in a denial of service, it does not allow privilege escalation, unauthorized data access, or code execution. Within Red Hat OpenShift AI and Red Hat Enterprise Linux AI, vLLM is delivered as container images supporting model inference, where the flaw affects multimodal instances that accept audio content. Deployments restricted to text-only language models or isolated behind authenticated gateways mitigate the likelihood and operational reach of this flaw. The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.