All vulnerabilities

Servers and applications

GHSA-mcmc-2m55-j8jj

Missing validation of multimodal embeddings leading to DoS and potential RCE

Severity
High 8.8
Published
8 Jan 2026
Last changed
8 Jan 2026
Added to UpHeron
3 Nov 2026
Fixed in
0.11.1
Weakness
Improper input validation (CWE-20)CWE-123Deserialization of untrusted data (CWE-502)Out-of-bounds write (CWE-787)
Attack
Over the network · Needs a user account · No user action needed
Attacker can
Read data · Change data · Take it down
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Sources
Project advisories (GitHub)

Description

The fix here for CVE-2025-62164 is not sufficient. The fix only disables prompt embeds by default rather than addressing the root cause, so the DoS vulnerability remains when the feature is enabled.

Our pending change attempts to fix the root cause, which is the missing sparse tensor validation. PyTorch (~v2.0) disables sparse tensor validation (specifically, sparse tensor invariants checks) by default for performance reasons. We are adding the sparse tensor validation to ensure indices are valid, non-negative, and within bounds. These checks help catch malformed tensors.

Current fix only added a flag to disable/enable prompt embeds, so by default, prompt embeds feature is disabled in vLLM, which stops DoS attacks through the embeddings. However, It doesn’t address the problem when the flag is enabled and there is still potential for DoS attacks.

From vllm-project/vllm's security advisory.

Affected software

SoftwareAffected versions
vLLMServers and applications · github:vllm-project/vllm
  • From 0.10.2 before 0.11.1

About vLLM

open-source software for large language model inference

From Wikidata, CC0 1.0.
Docker Hub pulls
44,195,225Docker Hub, 15 Nov 2026
GitHub stars
96,943GitHub, 15 Nov 2026
Forks
24,836GitHub, 15 Nov 2026
Latest release
v0.33.0rc3GitHub, 15 Nov 2026
Released
2 Nov 2026GitHub
License
Apache Software License 2.0Wikidata
First published on GitHub
9 Feb 2023GitHub

References