CVE-2025-71439
Published:October 27, 2026
Updated:October 28, 2026
llama.cpp before b5649 contains a heap out-of-bounds read vulnerability in llama_vocab::impl::print_info() that allows attackers to crash model loading via oversized special token IDs. Attackers can supply a crafted GGUF model with a short tokenizer.ggml.tokens array, causing llama-cli or llama-server to read past id_to_token and crash.
Affected Packages
llama-cpp (CONAN):
Affected version(s) >=b2038 <b6565Fix Suggestion:
Update to version b6565llama.cpp (CONDA):
Affected version(s) =5289 <5649Fix Suggestion:
Update to version 5649https://github.com/ggml-org/llama.cpp.git (GITHUB):
Affected version(s) >=b1046 <b5649Fix Suggestion:
Update to version b5649Related Resources (6)
Do you need more information?
Contact UsCVSS v4
Base Score:
6.8
Attack Vector
LOCAL
Attack Complexity
LOW
Attack Requirements
NONE
Privileges Required
NONE
User Interaction
PASSIVE
Vulnerable System Confidentiality
NONE
Vulnerable System Integrity
NONE
Vulnerable System Availability
HIGH
Subsequent System Confidentiality
NONE
Subsequent System Integrity
NONE
Subsequent System Availability
NONE
CVSS v3
Base Score:
5.5
Attack Vector
LOCAL
Attack Complexity
LOW
Privileges Required
NONE
User Interaction
REQUIRED
Scope
UNCHANGED
Confidentiality
NONE
Integrity
NONE
Availability
HIGH
Weakness Type (CWE)
Out-of-bounds Read
EPSS
Base Score:
0.12