hoch6.5
CVE-2025-62372
vLLM vulnerable to DoS with incorrect shape of multimodal embedding inputs
Auswirkung
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.
Ein Angreifer kann die Lücke über das Netzwerk ausnutzen, mit einem normalen Benutzerkonto, ohne Zutun eines Nutzers. Bei Erfolg: hohes Risiko für Ausfälle.
Betroffene Pakete
| Ökosystem | Paket | Betroffen | Behoben in |
|---|---|---|---|
| pip | vllm | >= 0.5.5, < 0.11.1 | 0.11.1 |
Betroffene Produkte · Vllm
| Produkt | Betroffene Versionen |
|---|---|
| Vllm | ab 0.5.5 vor 0.11.1 Version 0.11.1 |
Schwachstellenklasse
Referenzen
- NVD-Eintrag
- CVE.org-Eintrag
- GitHub-Advisory
- https://github.com/vllm-project/vllm/commit/58fab50d82838d5014f4a14d991fdb9352c9c84b
- https://github.com/vllm-project/vllm/pull/27204
- https://github.com/vllm-project/vllm/pull/6613
- https://github.com/vllm-project/vllm/security/advisories/GHSA-pmqf-x6x8-p7qw
- https://github.com/advisories/GHSA-pmqf-x6x8-p7qw
- https://nvd.nist.gov/vuln/detail/CVE-2025-62372
- https://github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2026-2019.yaml
- https://github.com/vllm-project/vllm
- https://pypi.org/project/vllm