Vulnerability · Published Aug 25, 2026 · Assigned by VulnCheck

CVE-2026-78684vLLM before 0.27.0 Denial of Service via DeepStream Backend

Medium 5.3

In plain English

A medium-severity uncontrolled resource consumption in vllm from vllm-project. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.6% (higher than 49% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.

Severity (CVSS 3.1)5.3Medium

Exploit odds (30 days)0.6%top 50.7% of all CVEs

Exploited in the wildNo reportsCISA SSVC: none

Attack toolingNoneno public tooling known

Official description

vLLM before 0.27.0 fails to properly classify DeepStream as a GPU backend and omits pixel-limit enforcement in its decode path. Unauthenticated attackers can activate DeepStream at request time to initialize the process-wide GPU decode pool and submit video that bypasses resource controls, causing partial denial of service for concurrent requests.

Affected packages & fixed versions

Ecosystem	Package	Fixed in	Advisory
PyPI	vllm	0.27.0	PYSEC-2026-4003

Affected products (as reported)

Vendor	Product	Versions
vllm-project	vllm	0 – before 0.27.0

Weakness: CWE-400 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L

References

GitHub Security Advisory (GHSA-cqm8-jxg6-fqfq) vendor-advisory
Patch Commit patch
VulnCheck Advisory: vLLM before 0.27.0 Denial of Service via DeepStream Backend third-party-advisory