Vulnerability · Published Sep 18, 2026 · Assigned by VulnCheck

CVE-2026-93592vLLM before 0.28.0 Denial of Service via negative token ID

High 7.5

In plain English

A high-severity security flaw in vllm from vllm-project. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 0.5% (higher than 46% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.

Severity (CVSS 3.1)7.5High

Exploit odds (30 days)0.5%top 54.1% of all CVEs

Exploited in the wildNo reportsCISA SSVC: poc

Attack toolingNoneno public tooling known

Official description

vLLM versions before 0.28.0 fail to validate the lower bound of token IDs in the /v1/embeddings and /pooling endpoints, allowing unauthenticated attackers to crash the engine by submitting negative token IDs. A single request with a negative token ID triggers a CUDA device-side assertion that poisons the GPU context, causing all subsequent requests to fail until the process restarts.

Affected packages & fixed versions

Ecosystem	Package	Fixed in	Advisory
PyPI	vllm	0.28.0	PYSEC-2026-3997

Affected products (as reported)

Vendor	Product	Versions
vllm-project	vllm	0 – before 0.28.0

Weakness: CWE-129 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

References

GitHub Security Advisory (GHSA-25q3-v2hm-8vpf) vendor-advisory
VulnCheck Advisory: vLLM before 0.28.0 Denial of Service via negative token ID third-party-advisory