PT-2026-27272 · Ggml Org · Llama.Cpp

Published 2026-03-24 · Updated 2026-03-24 · CVE-2026-33298

Mitre

CVSS v3.1

7.8

High

AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

llama.cpp is an inference of several LLM models in C/C++. Prior to b7824, an integer overflow vulnerability in the ggml nbytes function allows an attacker to bypass memory validation by crafting a GGUF file with specific tensor dimensions. This causes ggml nbytes to return a significantly smaller size than required (e.g., 4MB instead of Exabytes), leading to a heap-based buffer overflow when the application subsequently processes the tensor. This vulnerability allows potential Remote Code Execution (RCE) via memory corruption. b7824 contains a fix.

Fix

Integer Overflow

Heap Based Buffer Overflow

Weakness Enumeration

CWE-190

CWE-122

Related Identifiers

CVE-2026-33298

Affected Products

Llama.Cpp

References · 3

• https://nvd.nist.gov/vuln/detail/CVE-2026-33298 · Security Note
• https://github.com/ggml-org/llama.cpp/security/advisories/GHSA-96jg-mvhq-q7q7 · Note
• https://github.com/ggml-org/llama.cpp/releases/tag/b7824 · Note