llama.cpp is an inference of several LLM models in C/C++. Prior to version b5662, an attacker‐supplied GGUF model vocabulary can trigger a buffer overflow in llama.cpp’s vocabulary‐loading code. Specifically, the helper trycopy in llama.cpp/src/vocab.cpp: llamavocab::impl::tokentopiece() casts a very large sizet token length into an int32t, causing the length check (if (length < (int32t)size)) to be bypassed. As a result, memcpy is still called with that oversized size, letting a malicious model overwrite memory beyond the intended buffer. This can lead to arbitrary memory corruption and potential code execution. This issue has been patched in version b5662.
{
"cwe_ids": [
"CWE-119",
"CWE-195"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2025/49xxx/CVE-2025-49847.json",
"cna_assigner": "GitHub_M"
}{
"cpe": "cpe:2.3:a:ggml:llama.cpp:*:*:*:*:*:*:*:*",
"extracted_events": [
{
"introduced": "0"
},
{
"fixed": "b5662"
}
],
"source": [
"CPE_RANGE",
"REFERENCES"
]
}