Loading HuntDB...
Hunt
Vulnerability Intelligence by wss.sh
Home Daily Briefings High Impact CVEs Recent Exploits KEV Catalog GitHub Advisories Latest Updates Security News
Sign In Sign Up

CVE-2025-32444

CRITICAL
Published 2025-04-30T00:25:00.655Z
Actions:

Expert Analysis

Professional remediation guidance

Get tailored security recommendations from our analyst team for CVE-2025-32444. We'll provide specific mitigation strategies based on your environment and risk profile.

CVSS Score

V3.1
10.0
/10
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H
Base Score Metrics
Exploitability: N/A Impact: N/A

EPSS Score

v2025.03.14
0.007
probability
of exploitation in the wild

There is a 0.7% chance that this vulnerability will be exploited in the wild within the next 30 days.

Updated: 2025-06-25
Exploit Probability
Percentile: 0.718
Higher than 71.8% of all CVEs

Attack Vector Metrics

Attack Vector
NETWORK
Attack Complexity
LOW
Privileges Required
NONE
User Interaction
NONE
Scope
CHANGED

Impact Metrics

Confidentiality
HIGH
Integrity
HIGH
Availability
HIGH

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.

Available Exploits

No exploits available for this CVE.

Related News

CVE-2025-32444 (CVSS 10): Critical RCE Flaw in vLLM’s Mooncake Integration Exposes AI Infrastructure

A critical security vulnerability has been disclosed in vLLM, a popular open-source library used for high-performance inference and The post CVE-2025-32444 (CVSS 10): Critical RCE Flaw in vLLM’s Mooncake Integration Exposes AI Infrastructure appeared first on Daily CyberSecurity.

SecurityOnline.info 2025-05-01 00:35
Read More

Affected Products

vllm-project

vllm

Affected Versions:

>= 0.6.5, < 0.8.5

EU Vulnerability Database

Monitored by ENISA for EU cybersecurity

EU Coordination

EU Coordinated

Exploitation Status

No Known Exploitation

EUVD ID

View on ENISA

ENISA Analysis

vLLM Vulnerable to Remote Code Execution via Mooncake Integration

Affected Products (ENISA)

vllm-project
vllm

ENISA Scoring

CVSS Score (3.1)

10.0
/10
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H

EPSS Score

0.770
probability

ENISA References

https://github.com/vllm-project/vllm/security/advisories/GHSA-hj4w-hm2g-p6w5
https://github.com/vllm-project/vllm/security/advisories/GHSA-x3m8-f7g5-qhm7
https://github.com/vllm-project/vllm/commit/a5450f11c95847cf51a17207af9a3ca5ab569b2c
https://github.com/vllm-project/vllm/blob/32b14baf8a1f7195ca09484de3008063569b43c5/vllm/distributed/kv_transfer/kv_pipe/mooncake_pipe.py#L179
https://nvd.nist.gov/vuln/detail/CVE-2025-32444
https://github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2025-42.yaml

Data provided by ENISA EU Vulnerability Database. Last updated: May 29, 2025

GitHub Security Advisories

Community-driven vulnerability intelligence from GitHub

✓ GitHub Reviewed CRITICAL

vLLM Vulnerable to Remote Code Execution via Mooncake Integration

GHSA-hj4w-hm2g-p6w5

Advisory Details

## Impacted Deployments **Note that vLLM instances that do NOT make use of the mooncake integration are NOT vulnerable.** ## Description vLLM integration with mooncake is vaulnerable 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. This is a similar to [GHSA - x3m8 - f7g5 - qhm7](https://github.com/vllm-project/vllm/security/advisories/GHSA-x3m8-f7g5-qhm7), the problem is in https://github.com/vllm-project/vllm/blob/32b14baf8a1f7195ca09484de3008063569b43c5/vllm/distributed/kv_transfer/kv_pipe/mooncake_pipe.py#L179 Here [recv_pyobj()](https://github.com/zeromq/pyzmq/blob/453f00c5645a3bea40d79f53aa8c47d85038dc2d/zmq/sugar/socket.py#L961) Contains implicit `pickle.loads()`, which leads to potential RCE.

Affected Packages

PyPI vllm
ECOSYSTEM: ≥0.6.5 <0.8.5

CVSS Scoring

CVSS Score

9.0

CVSS Vector

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

References

WEB https://github.com/vllm-project/vllm/security/advisories/GHSA-hj4w-hm2g-p6w5
WEB https://github.com/vllm-project/vllm/security/advisories/GHSA-x3m8-f7g5-qhm7
ADVISORY https://nvd.nist.gov/vuln/detail/CVE-2025-32444
WEB https://github.com/vllm-project/vllm/commit/a5450f11c95847cf51a17207af9a3ca5ab569b2c
WEB https://github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2025-42.yaml
PACKAGE https://github.com/vllm-project/vllm
WEB https://github.com/vllm-project/vllm/blob/32b14baf8a1f7195ca09484de3008063569b43c5/vllm/distributed/kv_transfer/kv_pipe/mooncake_pipe.py#L179

Advisory provided by GitHub Security Advisory Database. Published: April 29, 2025, Modified: May 29, 2025

References

https://github.com/vllm-project/vllm/security/advisories/GHSA-hj4w-hm2g-p6w5
https://github.com/vllm-project/vllm/security/advisories/GHSA-x3m8-f7g5-qhm7
https://github.com/vllm-project/vllm/commit/a5450f11c95847cf51a17207af9a3ca5ab569b2c
https://github.com/vllm-project/vllm/blob/32b14baf8a1f7195ca09484de3008063569b43c5/vllm/distributed/kv_transfer/kv_pipe/mooncake_pipe.py#L179
Published: 2025-04-30T00:25:00.655Z
Last Modified: 2025-04-30T13:08:35.928Z
Copied to clipboard!