CVE WATCH / VULNERABILITY DETAIL
CVE-2026-115912
HIGHCVSS 7.5NVD feed
Published 26 October 2026 · tracked since 27 October 2026
Description
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, an authenticated caller can place forged shared-memory address and monotonic_id values in public multimodal features.kwargs_data. In scale-out deployments with shared-memory multimodal caching enabled, workers trust
References
- NVD — National Vulnerability Database
- CVE.org record (MITRE)
- CISA Known Exploited Vulnerabilities catalog
Latest tracked vulnerabilities
- CVE-2026-72752 MEDIUM 6.5
- CVE-2026-50095 MEDIUM 6.1
- CVE-2026-116398 MEDIUM 6.5
- CVE-2026-116400 MEDIUM 6.5
- CVE-2026-115922 MEDIUM 6.2
- CVE-2026-115923 MEDIUM 6.5
- CVE-2026-115966 MEDIUM 6.5
- CVE-2026-115918 MEDIUM 6.5
- CVE-2026-115919 MEDIUM 6.2
- CVE-2026-115920 MEDIUM 6.2