CVE-2026-118375: ggml-org llama.cpp (Critical, CVSS 9.2)
- CVSS
- 9.2 Critical (v4.0)
- EPSS
- 0.4% chance of exploitation in the next 30 days (higher than 30.5% of CVEs)
- Exploited in the wild
- Not on CISA’s KEV list
- Public exploit
- None reported in public exploit databases
- Weakness
- CWE-787
- Vector
CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N- Published
- Oct 28, 2026
What it is
llama.cpp b8841 before b11865 contains an out-of-bounds write vulnerability in the ggml-rpc RDMA transport rdma_recv() function that allows unauthenticated peers to overflow the rpc-server stack. Attackers advertising RDMA capabilities in the HELLO message can send up to 256 KiB into a 1-byte stack variable in rpc_serve_client(), crashing the server or potentially executing code.
Description from the National Vulnerability Database.
What to do
Check whether you run ggml-org llama.cpp, then apply the vendor’s fix or mitigation from the advisories below.
Sources: NVD, FIRST EPSS, CISA KEV, and public exploit databases (availability only). ThreatRoundup reports whether public exploit code exists but never links to it: fake proof-of-concepts that infect security researchers are common. This page updates automatically.