MEDIUM 6.3 NVD
CVE-2026-111392
SGLang through 0.5.21 trusts caller-supplied mm_hashes values as the multimodal prefix-cache key, allowing remote attackers to poison cached KV entries across r
SGLang through 0.5.21 trusts caller-supplied mm_hashes values as the multimodal prefix-cache key, allowing remote attackers to poison cached KV entries across requests. Attackers can send a different image under a known or predicted hash via /generate to disclose another request's image content or corrupt victims' responses.
References
- https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/managers/io_struct.py
- https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/managers/tokenizer_ma
- https://github.com/sgl-project/sglang/issues/44641
- https://github.com/sgl-project/sglang/pull/44653
- https://www.vulncheck.com/advisories/sglang-through-0.5.21-kv-cache-poisoning-via-mm-hashe
This medium severity vulnerability with a CVSS score of 6.3 was published on 2026-10-16 via NVD.
vulnfeed aggregates 16467 vulnerabilities from NVD, CISA KEV,
Ubuntu, Debian, Red Hat, Kubernetes, Exploit-DB, OSS-Security, GitHub and OpenStack — updated every 4 hours.