---
id: CVE-2026-116180
title: >-
  vLLM from 0.8.5 before 0.31.0 Cross-Tenant DoS via Uninitialized
  LogprobsTensors
summary: >-
  vLLM from 0.8.5 before 0.31.0 has a denial of service flaw.
  LogprobsTensors.empty_cpu() in vllm/v1/outputs.py allocates an uninitialized
  int32 tensor, and the V1 model runner leaves prompt positions reported as
  computed by a KV connector…
severity: high
cvss: 7.7
cvssVector: 'CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:N/I:N/A:H'
cvssSource: cna
cwe:
  - CWE-908
vendor: vllm-project
product: vllm
affected:
  - vllm >= 0.8.5 < 0.31.0
ssvc:
  exploitation: none
  automatable: 'no'
  technicalImpact: partial
  timestamp: '2026-10-26T19:38:04.390533Z'
published: '2026-10-25'
updated: '2026-10-26'
sourceUpdated: '2026-10-26T19:48:35.789Z'
source: CVEORG
sourceUrl: 'https://www.cve.org/CVERecord?id=CVE-2026-116180'
references:
  - url: >-
      https://github.com/vllm-project/vllm/security/advisories/GHSA-rhcx-5729-88vg
    label: GitHub Security Advisory (GHSA-rhcx-5729-88vg)
  - url: >-
      https://www.vulncheck.com/advisories/vllm-from-0.8.5-before-0.31.0-cross-tenant-dos-via-uninitialized-logprobstensors
    label: >-
      VulnCheck Advisory: vLLM from 0.8.5 before 0.31.0 Cross-Tenant DoS via
      Uninitialized LogprobsTensors
  - url: >-
      https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-116180.json
  - url: 'https://access.redhat.com/security/cve/CVE-2026-116180'
  - url: 'https://bugzilla.redhat.com/show_bug.cgi?id=2560429'
  - url: 'https://www.cve.org/CVERecord?id=CVE-2026-116180'
  - url: 'https://nvd.nist.gov/vuln/detail/CVE-2026-116180'
tags:
  - cve.org
  - csaf
  - vex
  - red-hat
epss: 0.00249
epssPercentile: 0.16108
scores:
  cna: 7.7
  vendor: 6.5
ingestedAt: '2026-10-25T22:08:29.945Z'
---

## Overview

vLLM from 0.8.5 before 0.31.0 has a denial of service flaw. LogprobsTensors.empty_cpu() in vllm/v1/outputs.py allocates an uninitialized int32 tensor, and the V1 model runner leaves prompt positions reported as computed by a KV connector unwritten. A client sharing the engine can send prompt_logprobs (or echo with logprobs) against a connector-cached prefix, making tokenizer decode raise OverflowError, which crashes EngineCore and drops all co-tenant requests.

## Affected

- `vllm >= 0.8.5 < 0.31.0`

## Remediation

Refer to the linked advisories for vendor-supplied fixes and affected version ranges.

## Vendor advisories

- **Red Hat VEX** · Moderate · affected: Red Hat AI Inference Server, Red Hat Enterprise Linux AI (RHEL AI) 3, Red Hat OpenShift AI (RHOAI) · no fix planned: Red Hat OpenShift AI (RHOAI), Red Hat AI Inference Server, Red Hat Enterprise Linux AI (RHEL AI) 3 · updated 2026-10-26 · [vex](https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-116180.json)