CVE-2026-22773
January 10, 2026
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.
Affected Packages
vllm (CONDA):
Affected version(s) >=0.8.3 <0.12.0Fix Suggestion:
Update to version 0.12.0https://github.com/vllm-project/vllm.git (GITHUB):
Affected version(s) >=v0.1.0 <v0.12.0Fix Suggestion:
Update to version v0.12.0vllm (PYTHON):
Affected version(s) >=0.6.4 <0.12.0Fix Suggestion:
Update to version 0.12.0vllm (PYTHON):
Affected version(s) >=0.0.1 <0.12.0Fix Suggestion:
Update to version 0.12.0Related Resources (5)
Do you need more information?
Contact UsCVSS v4
Base Score:
7.1
Attack Vector
NETWORK
Attack Complexity
LOW
Attack Requirements
NONE
Privileges Required
LOW
User Interaction
NONE
Vulnerable System Confidentiality
NONE
Vulnerable System Integrity
NONE
Vulnerable System Availability
HIGH
Subsequent System Confidentiality
NONE
Subsequent System Integrity
NONE
Subsequent System Availability
NONE
CVSS v3
Base Score:
6.5
Attack Vector
NETWORK
Attack Complexity
LOW
Privileges Required
LOW
User Interaction
NONE
Scope
UNCHANGED
Confidentiality
NONE
Integrity
NONE
Availability
HIGH
Weakness Type (CWE)
Allocation of Resources Without Limits or Throttling
EPSS
Base Score:
0.05
