Home
Home
Trends
Trends
Vulnerabilities
Vulnerabilities
News
News
Researchers
Researchers
Why dbugs?
Why dbugs?
Settings

Oxcabe

#42947of 60,753
7.5Total CVSS
Vulnerabilities · 1
PT-2026-2260
2026-01-10
7.5
CVE-2026-22773
Vllm
Vllm
vLLM is an inference and serving engine for large language models (LLMs). Users can cause the vLLM engine to crash when serving multimodal models that utilize the Idefics3 vision model implementation. This is achieved by submitting a specifically designed 1x1 pixel image. The crafted image triggers a tensor dimension mismatch, resulting in an unhandled runtime error and complete server termination.