Description
vLLM is an inference and serving engine for large language models (LLMs). In versions from 0.6.4 to before 0.12.0, users can crash the vLLM engine serving multimodal models that use the Idefics3 vision model implementation by sending a specially crafted 1x1 pixel image. This causes a tensor dimension mismatch that results in an unhandled runtime error, leading to complete server termination. This issue has been patched in version 0.12.0.
CVSS Metrics
- Vector
- CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
- Attack Vector
- network
- Complexity
- low
- Privileges
- low
- User Action
- none
- Scope
- unchanged
- Confidentiality
- none
- Integrity
- none
- Availability
- high
- Weaknesses
- CWE-770
Metadata
- Primary Vendor
- VLLM
- Published
- 1/10/2026
- Last Modified
- 1/27/2026
- Source
- NIST NVD
- Note: Verify all details with official vendor sources before applying patches.
Affected Products
vllm : vllm
AI-Powered Remediation
Generate remediation guidance or a C-suite brief for this vulnerability.