Description
NVIDIA TensorRT-LLM for any platform contains a vulnerability in visual gen server, where an attacker could cause an unsafe deserialization by unauthorized zeroMQ deserialization. A successful exploit of this vulnerability might lead to code execution.
Published: 2026-07-14
Score: 6.4 Medium
EPSS: < 1% Very Low
KEV: No
Impact: n/a
Action: n/a
AI Analysis

Impact

A flaw in NVIDIA TensorRT‑LLM’s visual generation server permits the deserialization of messages received through ZeroMQ without proper validation. An attacker capable of sending crafted ZeroMQ payloads can exploit this unsafe deserialization to trigger arbitrary code execution, as the vulnerability is a classic instance of CWE‑502. The description states that a successful exploitation may lead to code execution, indicating that the impact is full compromise of the system hosting the server.

Affected Systems

The vulnerability affects NVIDIA TensorRT‑LLM on any platform. No specific version numbers are listed, so the risk applies to all releases of the software until an official fix is applied.

Risk and Exploitability

The CVSS score of 6.4 denotes a medium severity, and the EPSS score of less than 1% indicates a low probability of exploitation. The vulnerability is not listed in CISA’s KEV catalog. Based on the description, the likely attack vector is an unauthorized remote user who can reach the visual generation server and send a malicious ZeroMQ message, causing unsafe deserialization and potential code execution.

Generated by OpenCVE AI on July 31, 2026 at 05:02 UTC.

Remediation

No vendor fix or workaround currently provided.

OpenCVE Recommended Actions

  • Acquire and install the latest NVIDIA TensorRT‑LLM patch that addresses the deserialization flaw.
  • Restrict ZeroMQ traffic to the visual generation server by allowing only trusted IP addresses or enforcing VPN access.
  • Monitor server logs for unusual deserialization activity or signs of unexpected code execution, and investigate any anomalies promptly.

Generated by OpenCVE AI on July 31, 2026 at 05:02 UTC.

Tracking

Sign in to view the affected projects.

Advisories

No advisories yet.

History

Fri, 31 Jul 2026 05:30:00 +0000

Type Values Removed Values Added
Title ZeroMQ Unsafe Deserialization in NVIDIA TensorRT-LLM Visual Generation Server

Tue, 28 Jul 2026 05:00:00 +0000

Type Values Removed Values Added
Title Unsafe Deserialization via ZeroMQ in NVIDIA TensorRT-LLM Visual Generation Server

Sat, 25 Jul 2026 09:00:00 +0000

Type Values Removed Values Added
Title Unsafe Deserialization via ZeroMQ in NVIDIA TensorRT-LLM Visual Generation Server

Mon, 20 Jul 2026 21:45:00 +0000

Type Values Removed Values Added
Title Unauthorized ZeroMQ Deserialization in NVIDIA TensorRT-LLM

Fri, 17 Jul 2026 05:30:00 +0000

Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'no', 'Exploitation': 'none', 'Technical Impact': 'total'}, 'version': '2.0.3'}


Thu, 16 Jul 2026 13:45:00 +0000

Type Values Removed Values Added
Title Unauthorized ZeroMQ Deserialization in NVIDIA TensorRT-LLM

Wed, 15 Jul 2026 16:00:00 +0000

Type Values Removed Values Added
First Time appeared Nvidia
Nvidia tensorrt-llm
Vendors & Products Nvidia
Nvidia tensorrt-llm

Tue, 14 Jul 2026 20:30:00 +0000

Type Values Removed Values Added
Description NVIDIA TensorRT-LLM for any platform contains a vulnerability in visual gen server, where an attacker could cause an unsafe deserialization by unauthorized zeroMQ deserialization. A successful exploit of this vulnerability might lead to code execution.
Weaknesses CWE-502
References
Metrics cvssV3_1

{'score': 6.4, 'vector': 'CVSS:3.1/AV:L/AC:H/PR:H/UI:N/S:U/C:H/I:H/A:H'}


Subscriptions

Nvidia Tensorrt-llm
cve-icon MITRE

Status: PUBLISHED

Assigner: nvidia

Published:

Updated: 2026-07-15T14:19:44.134Z

Reserved: 2026-01-21T19:09:35.635Z

Link: CVE-2026-24220

cve-icon Vulnrichment

Updated: 2026-07-15T14:19:39.326Z

cve-icon NVD

No data.

cve-icon Redhat

No data.

cve-icon OpenCVE Enrichment

Updated: 2026-07-31T05:15:03Z

Weaknesses
  • CWE-502

    Deserialization of Untrusted Data