vLLM 0.6.30.9.0 ReDoS in LoRA & Chat Endpoints
CVE-2025-71379 Published on June 20, 2026

vllm - Regular Expression Denial of Service in Multiple Components
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.

Vendor Advisory NVD

Vulnerability Analysis

CVE-2025-71379 can be exploited with network access, and requires small amount of user privileges. This vulnerability is considered to have a low attack complexity. The potential impact of an exploit of this vulnerability is considered to have no impact on confidentiality and integrity, and a small impact on availability.

Attack Vector:
NETWORK
Attack Complexity:
LOW
Privileges Required:
LOW
User Interaction:
NONE
Scope:
UNCHANGED
Confidentiality Impact:
NONE
Integrity Impact:
NONE
Availability Impact:
LOW

Weakness Type

What is a ReDoS Vulnerability?

The product uses a regular expression with an inefficient, possibly exponential worst-case computational complexity that consumes excessive CPU cycles. Some regular expression engines have a feature called "backtracking". If the token cannot match, the engine "backtracks" to a position that may result in a different token that can match. Backtracking becomes a weakness if all of these conditions are met:

CVE-2025-71379 has been classified to as a ReDoS vulnerability or weakness.


Products Associated with CVE-2025-71379

Want to know whenever a new CVE is published for Vllm? stack.watch will email you.

 

Affected Versions

vllm: