vllm-project · vLLM · published 2026-06-17
CVE-2026-12491Vllm: vllm: image exif rotation & png trns transparency not normalized, causing mismatch between model input and expectations
Routine No exploitation evidence; EPSS 0.2%.
CVE-2026-12491 is not known to be exploited in the wild. EPSS estimates a 0.2% chance of exploitation in the next 30 days, higher than 15% of all CVEs. This vLLM vulnerability: CVSS 4.8 (v3.1). Fixed upstream in vLLM 0.24.0. Routine: patch when convenient.
Jump to: Remediation Exploit intelAffected versionsDescription
Remediation#
Fixed upstream in vLLM 0.24.0.
Patch events
AI toolingvLLM
vLLM security release: 0.24.0
Upgrade to
0.24.0Why: CVE-2026-54236: Public proof-of-concept exists (SSVC exploitation: poc).
Exploit intelligence#
| Signal | Value | Source |
|---|---|---|
| Known exploited | No | CISA KEV |
| Ransomware use | — | CISA KEV |
| EPSS (30-day exploitation probability) | 0.2% (15th pct) | FIRST EPSS v2026.06.15 |
| CVSS base score | 4.8 (v3.1)CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:L | CNA / CISA ADP |
| SSVC exploitation | none | CISA Vulnrichment |
| SSVC automatable | No | CISA Vulnrichment |
| SSVC technical impact | partial | CISA Vulnrichment |
Affected versions (from the CVE record)#
- vllm-project vLLM
0.11.0to before0.24.0(affected)
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
- Red Hat Red Hat AI Inference Server
Description (from the CVE record)#
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.
Quoted from the CVE Program record.