CVE-2025-48943

Published May 30, 2025

Last updated a month ago

CVSS medium 6.5

Overview

Description
vLLM is an inference and serving engine for large language models (LLMs). Version 0.8.0 up to but excluding 0.9.0 have a Denial of Service (ReDoS) that causes the vLLM server to crash if an invalid regex was provided while using structured output. This vulnerability is similar to GHSA-6qc9-v4r8-22xg/CVE-2025-48942, but for regex instead of a JSON schema. Version 0.9.0 fixes the issue.
Source
security-advisories@github.com
NVD status
Undergoing Analysis

Risk scores

CVSS 3.1

Type
Secondary
Base score
6.5
Impact score
3.6
Exploitability score
2.8
Vector string
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Severity
MEDIUM

Weaknesses

security-advisories@github.com
CWE-248

Social media

Hype score
Not currently trending

  1. CVE-2025-48943 vLLM is an inference and serving engine for large language models (LLMs). Version 0.8.0 up to but excluding 0.9.0 have a Denial of Service (ReDoS) that causes the vLL… https://t.co/Pzc0j3LZcD

    @CVEnew

    30 May 2025

    347 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

References

Sources include official advisories and independent security research.