---
id: CVE-2024-8939
aliases:
  - GHSA-wc36-9694-f9rf
  - PYSEC-2026-2025
title: vLLM Denial of Service via the best_of parameter
summary: vLLM Denial of Service via the best_of parameter
severity: medium
cvss: 6.2
cvssVector: 'CVSS:3.1/AV:L/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H'
vendor: vllm
product: vllm
ecosystem: pip
affected:
  - vllm <= 0.5.0.post1
published: '2024-09-17'
updated: '2026-07-07'
source: OSV
sourceUrl: 'https://osv.dev/vulnerability/GHSA-wc36-9694-f9rf'
references:
  - url: 'https://nvd.nist.gov/vuln/detail/CVE-2024-8939'
  - url: 'https://github.com/vllm-project/vllm/issues/6137'
  - url: 'https://access.redhat.com/security/cve/CVE-2024-8939'
  - url: 'https://bugzilla.redhat.com/show_bug.cgi?id=2312782'
  - url: 'https://github.com/vllm-project/vllm'
tags:
  - osv
  - pip
epss: 0.00233
epssPercentile: 0.12683
ingestedAt: '2026-07-08T18:25:53.597Z'
---

## Overview

A vulnerability was found in the ilab model serve component, where improper handling of the best_of parameter in the vllm JSON web API can lead to a Denial of Service (DoS). The API used for LLM-based sentence or chat completion accepts a best_of parameter to return the best completion from several options. When this parameter is set to a large value, the API does not handle timeouts or resource exhaustion properly, allowing an attacker to cause a DoS by consuming excessive system resources. This leads to the API becoming unresponsive, preventing legitimate users from accessing the service.

## Affected packages

- `vllm <= 0.5.0.post1`

## Remediation

Refer to the advisory for the patched release.