Chainggml-org · llama.cpp
ggml-org llama.cpp: fix bypasses (2025–2026)
3 CVEs over 2 generations, 10 Jul 2025 to 27 Oct 2026.
Family tree
Left to right by publication date.Oldest first. Select a CVE to open it.- CVE-2025-5363010 Jul 2025 First in the chain
- CVE-2026-2794012 Mar 2026 related to CVE-2025-53630
- CVE-2026-11684027 Oct 2026 bypassed the fix for CVE-2025-53630
Exploited in the wild (CISA KEV) Not in KEV Priority score (0 to 100) High confidence Medium Low Link confidenceHow strong the evidence is that two CVEs are related. High: stated in the CVE text or linked in an advisory, or the same code was patched one release apart. Medium: the same code, or a version step plus a similar description. Low: similar description only.
CVEs in this family
Table view of the tree| CVE | Published | Title | Linked from | Priority |
|---|---|---|---|---|
| CVE-2025-53630 | 10 Jul 2025 | Integer Overflow in GGUF Parser can lead to Heap Out-of-Bounds Read/Write in gguf | First in chain | 12.3 |
| CVE-2026-27940 | 12 Mar 2026 | llama.cpp has a Heap Buffer Overflow via Integer Overflow in `mem_size` Calculation — Bypass of CVE-2025-53630 Fix | CVE-2025-53630 · related · Medium | 12.1 |
| CVE-2026-116840 | 27 Oct 2026 | llama.cpp b5854 before b11275 Integer Overflow in gguf_init_from_reader | CVE-2025-53630 · bypassed the fix · High | 4.7 |