CVE-2026-12491

A flaw was found in vLLM, an open-source library for large language model inference. This vulnerability arises from improper handling of image metadata, specifically EXIF orientation and PNG transparency (tRNS) data, during image processing. When images are converted to RGB, transparency information may be implicitly discarded or remapped, leading to unexpected rendering of transparent pixels and distortion of input content. This can result in the model misinterpreting image content, potentially affecting the integrity of processed data.
Configurations

No configuration.

History

17 Jun 2026, 16:14

Type Values Removed Values Added
New CVE

Information

Published : 2026-06-17 13:20

Updated : 2026-07-07 09:16


NVD link : CVE-2026-12491

Mitre link : CVE-2026-12491

CVE.ORG link : CVE-2026-12491


JSON object : View

Products Affected

No product.

CWE
CWE-115

Misinterpretation of Input