CVE-2026-12491

published 17 Jun 2026 · updated 7 Jul 2026

Vllm: vllm: image exif rotation & png trns transparency not normalized, causing mismatch between model input and expectations

4.8MediumCVSS 4.8EPSS 0%

No exploitation in the wild has been reported. EPSS puts the chance of exploitation in the next 30 days at 0%. Fixed in 0.24.0.

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.

Affected

Vendors, products and packages with version ranges

VendorProductVersionsFixed
Red HatAI Inference Server
Red HatEnterprise Linux
Red HatOpenShift AI
vllm-project>=0.11.0 <0.24.00.24.0

Timeline

Every event on this vulnerability

Related

Other vulnerabilities in Red Hat, strongest first

Embed

A live severity badge for an advisory, README or ticket

CVE-2026-12491 severity on Zeptric[![CVE-2026-12491](https://zeptric.io/badge/cve/CVE-2026-12491.svg)](https://zeptric.io/cve/CVE-2026-12491)