Home

Description

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Versions starting from 0.6.5 and prior to 0.8.5, having vLLM integration with mooncake, are vulnerable to remote code execution due to using pickle based serialization over unsecured ZeroMQ sockets. The vulnerable sockets were set to listen on all network interfaces, increasing the likelihood that an attacker is able to reach the vulnerable ZeroMQ sockets to carry out an attack. vLLM instances that do not make use of the mooncake integration are not vulnerable. This issue has been patched in version 0.8.5.

PUBLISHED Reserved 2025-04-08 | Published 2025-04-30 | Updated 2025-04-30 | Assigner GitHub_M




CRITICAL: 10.0CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H

Problem types

CWE-502: Deserialization of Untrusted Data

Product status

>= 0.6.5, < 0.8.5
affected

References

github.com/...t/vllm/security/advisories/GHSA-hj4w-hm2g-p6w5

github.com/...t/vllm/security/advisories/GHSA-x3m8-f7g5-qhm7

github.com/...ommit/a5450f11c95847cf51a17207af9a3ca5ab569b2c

github.com/...stributed/kv_transfer/kv_pipe/mooncake_pipe.py

cve.org (CVE-2025-32444)

nvd.nist.gov (CVE-2025-32444)

Download JSON