CVE-2026-44222
Published: May 12, 2026
Modified: May 13, 2026
CVSS v3.1
6.5
Description
vLLM is an inference and serving engine for large language models (LLMs). From 0.6.1 to before 0.20.0, there is a a Token Injection vulnerability in vLLM’s multimodal processing. Unauthenticated, text-only prompts that spell special tokens are interpreted as control. Image and video placeholder sequences supplied without matching data cause vLLM to index into empty grids during input-position computation, raising an unhandled IndexError and terminating the worker or degrading availability. Multimodal paths that rely on image_grid_thw/video_grid_thw are affected. This vulnerability is fixed in 0.20.0.
| Vendor | Product | Versions |
|---|---|---|
vllm-project | vllm | affected >= 0.6.1, < 0.20.0 |
Weaknesses (CWE)
CVSS v3.1 Details
CVSS v3.1 Vector
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Attack Vector
Attack Complexity
Privileges Required
User Interaction
Scope
Confidentiality
Integrity
Availability
References
Security Training
Train your team to recognize and prevent security threats with our comprehensive security awareness program.
Start TrainingVulnerability Scanning
Discover vulnerabilities in your applications and infrastructure before attackers do.
Scan Now