CVE-2026-93592

vLLM before 0.28.0 Denial of Service via negative token ID

Published Sep 18, 2026 CVSS 4.0 · 8.7 High EPSS < 1% KEV · Not listed
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.

Analysis

AI-generated

Technical impact

The vulnerability exists in vLLM versions prior to 0.28.0, where the engine does not verify that token IDs sent to the /v1/embeddings or /pooling endpoints are non‑negative. An attacker who can access these public endpoints can supply a negative token ID, causing a CUDA device‑side assertion that corrupts the GPU context. Once triggered, all subsequent requests to the affected endpoints fail, effectively denying service until the process is restarted.

Affected Systems

Affected products are vllm-project’s vLLM component versions older than 0.28.0. Modern installations that have been updated to 0.28.0 or later are not impacted. The information might lack precise sub‑version ranges beyond the stated cutoff, so any deployment using a pre‑0.28.

Risk and Exploitability

The CVSS score of 8.7 reflects a high severity denial of service. The flaw can be exploited through unauthenticated HTTP requests to the endpoints, so privileged access is not required. The EPSS score of <1% indicates a very low probability of exploitation in the general population. The vulnerability is not listed in the CISA KEV catalog, yet it could still be leveraged in targeted attacks. No special network reachability is required beyond access to the service’s REST API.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Upgrade vLLM to version 0.28.0 or later, which includes validation for token IDs.
  • If an upgrade cannot be performed immediately, add input validation to reject any token ID less than zero before the request is processed.
  • After applying the fix or validation, restart the engine to recover from any poisoned GPU contexts.
  • Consider restricting access to the /v1/embeddings and /pooling endpoints to authenticated or internal users until a patch is available.

Generated by OpenCVE AI on September 23, 2026 at 01:42 UTC.

Tracking

Sign in to view the affected projects.

Advisories

No advisories available
History
Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'yes', 'Exploitation': 'poc', 'Technical Impact': 'partial'}, 'version': '2.0.3'}

Type Values Removed Values Added
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.
Title vLLM before 0.28.0 Denial of Service via negative token ID
First Time appeared Vllm
Vllm vllm
Weaknesses CWE-129
CPEs cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*
Vendors & Products Vllm
Vllm vllm
References
Metrics cvssV3_1

{'score': 7.5, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H'}

cvssV4_0

{'score': 8.7, 'vector': 'CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N'}

Weaknesses
  • CWE-129

    Improper Validation of Array Index

  • CWE-617

    Reachable Assertion

Sources

Status
PUBLISHED
Assigner
VulnCheck
Published
Updated
Sep 18, 2026, 6:09 PM
Reserved
Sep 18, 2026, 10:59 AM
Status
Analyzed
Published
Sep 18, 2026, 2:19 PM
Modified
Sep 28, 2026, 6:37 PM
Severity
Important
Public date
Sep 18, 2026, 1:20 PM