CVE-2026-93436

vLLM through 0.29.0 Memory Exhaustion via Rejected Requests

Published Sep 17, 2026 CVSS 4.0 · 8.7 High EPSS < 1% KEV · Not listed
Description
vLLM through 0.29.0 fails to properly clean up decode-side metadata for rejected inference requests in prefill/decode disaggregated deployments. Remote attackers can submit requests with max_tokens=0 to exhaust decode-worker memory without bound until the worker restarts.

Analysis

AI-generated

Technical impact

vLLM through 0.29.0 contains a memory leak due to improper cleanup of metadata for rejected inference requests. Attackers can submit requests with max_tokens=0 to force continual allocation of memory on decode workers until the service restarts, resulting in denial of service. The weakness is a classic memory exhaustion vulnerability (CWE-401 and CWE-770).

Affected Systems

The vulnerability affects all vLLM releases up to and including 0.29.0, running on any platform supported by the project (Linux, Windows). Users employing these versions should verify that their deployment uses the latest release.

Risk and Exploitability

The CVSS score of 8.7 reflects a high severity impact, but the EPSS score of less than 1% indicates a very low probability of exploitation at present. The exploit does not provide arbitrary code execution; it requires remote submission of specific request payloads to an exposed vLLM endpoint and is a resource‑based attack that forces worker restarts. Because the attack is purely denial of service and hinges on a small, predictable payload, it remains manageable through patching or defensive filtering.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Apply the latest vLLM release that includes the memory‑cleanup fix, at minimum v0.29.1 or newer.
  • If immediate upgrade is not possible, configure the application to reject or deny any inference request where max_tokens equals zero before it reaches the decode worker.
  • Implement runtime monitoring of memory usage for decode workers and set alerts to trigger if memory consumption exceeds expected thresholds, ensuring prompt remediation if the issue resurfaces.

Generated by OpenCVE AI on September 23, 2026 at 01:43 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': 'none', 'Technical Impact': 'partial'}, 'version': '2.0.3'}

Type Values Removed Values Added
Description vLLM through 0.29.0 fails to properly clean up decode-side metadata for rejected inference requests in prefill/decode disaggregated deployments. Remote attackers can submit requests with max_tokens=0 to exhaust decode-worker memory without bound until the worker restarts.
Title vLLM through 0.29.0 Memory Exhaustion via Rejected Requests
First Time appeared Vllm
Vllm vllm
Weaknesses CWE-401
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-401

    Missing Release of Memory after Effective Lifetime

  • CWE-770

    Allocation of Resources Without Limits or Throttling

Sources

Status
PUBLISHED
Assigner
VulnCheck
Published
Updated
Sep 24, 2026, 2:23 PM
Reserved
Sep 17, 2026, 9:50 PM
Status
Analyzed
Published
Sep 17, 2026, 11:18 PM
Modified
Sep 28, 2026, 6:37 PM
Severity
Important
Public date
Sep 17, 2026, 10:15 PM