CVE-2025-62372
vLLM is an inference and serving engine for large language models (LLMs). From version 0.5.5 to before 0.11.1, users can crash the vLLM engine serving multimodal models by passing multimodal embedding inputs with correct ndim but incorrect shape (e.g. hidden dimension is wrong), regardless of whether the model is intended to support such inputs (as defined in the Supported Models page). This issue has been patched in version 0.11.1.
Scoring
- Severity
- HIGH
- CVSS base score
- 8.3
- CVSS vector
- CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:H
- EPSS probability
- 0.38%
- CWE
- CWE-129
- Published
- 2025-11-21
- Last modified
- 2026-03-12
Affected products
- vllm-project vllm
Weakness type
Related vulnerabilities
- CVE-2026-49838 — GoBGP confederation validation panics on empty AS_PATH attribute
- CVE-2026-88052 — Tesseract: Heap out-of-bounds write in UNICHARSET::load_via_fgets via count/insert desynchronization
- CVE-2023-54396 — PocketMine-MP before 4.8.1 Server Crash via Banner NBT
- CVE-2026-87500 — Improper validation of array index in ANGLE in Google Chrome prior to 153.0.8010.36 allowed a...
- CVE-2026-0799 — OOBR and OOBW in libpcap before 1.10.7
- CVE-2026-57159 — PJSIP: SDP parser out-of-bounds write in remote payload-type map maintenance
- CVE-2026-85084 — Out-of-bounds write in TizenFX MediaBufferBase indexer setter due to missing bounds check
- CVE-2023-31308 — A malicious virtual function can invoke the certain command handlers in the SMU, causing a denial...