CVE-2025-46722

Published May 29, 2025

Last updated 2 months ago

Overview

Description
vLLM is an inference and serving engine for large language models (LLMs). In versions starting from 0.7.0 to before 0.9.0, in the file vllm/multimodal/hasher.py, the MultiModalHasher class has a security and data integrity issue in its image hashing method. Currently, it serializes PIL.Image.Image objects using only obj.tobytes(), which returns only the raw pixel data, without including metadata such as the image’s shape (width, height, mode). As a result, two images of different sizes (e.g., 30x100 and 100x30) with the same pixel byte sequence could generate the same hash value. This may lead to hash collisions, incorrect cache hits, and even data leakage or security risks. This issue has been patched in version 0.9.0.
Source
security-advisories@github.com
NVD status
Awaiting Analysis

Risk scores

CVSS 3.1

Type
Secondary
Base score
4.2
Impact score
2.5
Exploitability score
1.6
Vector string
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:L
Severity
MEDIUM

Weaknesses

security-advisories@github.com
CWE-1023

Social media

Hype score
Not currently trending