Vulnerability in Vllm
CVE-2025-71379
vLLM versions >= 0.6.3 and < 0.9.0 contain multiple regular expression denial of service (ReDoS) vulnerabilities. Several regex patterns — in vllm/lora/utils.py, the phi4mini tool parser, and the OpenAI-compatible serving chat endpoint — are susceptible to catastrophic backtracking. An attacker submitting crafted input with nested or repeated structures can trigger severe CPU consumption and performance degradation, resulting in denial of service.
Vulnerability class: ReDoS (Regular Expression Denial of Service)
Published · last modified .
CVSS v3 metric
CVSS v3 base score 4.3 (Medium). Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L.
EPSS exploit prediction
EPSS: 0.005 (39.0th percentile), scored .
Very low probability of exploitation in the next 30 days; routine patching cadence is appropriate. 39th percentile — 39.0% of CVEs in the catalogue have a lower EPSS than this one. How to read EPSS.
EPSS over last 30 days · oldest: 0.005 · newest: 0.005 · change: 0.000
Affected products
- Vllm — versions 0.6.3, 0.9.0
Weakness classification (CWE)
References
- github.com/vllm-project/vllm/security/advisories/GHSA-j828-28rj-hfhp (Exploit, vendor-advisory, Vendor Advisory)
- www.vulncheck.com/advisories/vllm-regular-expression-denial-of-service-in-multi… (Third Party Advisory, third-party-advisory)
Frequently asked questions
- What is CVE-2025-71379?
- CVE-2025-71379 is a medium-severity vulnerability in Vllm, classified under Inefficient Regular Expression Complexity. CVSS score: 4.3/10. Published 2026-06-20.
- How severe is CVE-2025-71379?
- Medium severity. CVSS v3 base score is 4.3 out of 10.