推理栈新增攻击面,恶意模型文件本身即触发条件,CVSS 7.0。
rcellmcve
llama.cpp MTMD Multimodal GGUF File Parsing Integer Overflow Remote Code Execution Vulnerability
November 13th, 2026
ZDI-26-954
ZDI-CAN-30980
CVE ID
CVE-2026-126976
CVSS Score
7.0
AV:L/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:H
Affected Vendors
llama.cpp
Affected Products
llama.cpp
Vulnerability Details
This vulnerability allows remote attackers to execute arbitrary code on affected installations of llama.cpp. User interaction is required to exploit this vulnerability in that the target must open a malicious file.
The specific flaw exists within the parsing of audio multimodal projector GGUF files. The issue results from the lack of proper validation of user-supplied data, which can result in an integer overflow before allocating a buffer. An attacker can leverage this vulnerability to execute arbitrary code in the context of the current process.
Additional Details
llama.cpp has issued an update to correct this vulnerability. More details can be found at:
https://github.com/ggml-org/llama.cpp/commit/e2e7a9b2d04e0404131e551243d4bd399ccfe606
Disclosure Timeline
2026-06-18
- Vulnerability reported to vendor
2026-11-13
- Coordinated public release of advisory
2026-11-13
- Advisory Updated
Credit
zent
Back to Advisories