Denial of Service Vulnerability in vLLM's Inference Engine
CVE-2025-48943

6.5MEDIUM

Key Information:

Vendor

Vllm-project

Status
Vllm
Vendor
CVE Published:
30 May 2025

What is CVE-2025-48943?

The vLLM inference and serving engine for large language models has a vulnerability that allows for Denial of Service (ReDoS) under specific conditions. When an invalid regular expression is provided while utilizing structured output, the vLLM server may crash, disrupting operations. This issue is present in versions from 0.8.0 up to, but not including, 0.9.0. Users are advised to update to the latest version to ensure protection against this vulnerability.

Affected Version(s)

vllm >= 0.8.0, < 0.9.0

References

https://github.com/vllm-project/vllm/security/advisories/...
x_refsource_CONFIRM
https://github.com/vllm-project/vllm/issues/17313
x_refsource_MISC
https://github.com/vllm-project/vllm/pull/17623
x_refsource_MISC

CVSS V3.1

Score:
6.5
Severity:
MEDIUM
Confidentiality:
None
Integrity:
None
Availability:
None
Attack Vector:
Network
Attack Complexity:
Low
Privileges Required:
Low
User Interaction:
None
Scope:
Unchanged

Timeline

  • Vulnerability published

  • Vulnerability Reserved

.