Description
llama.cpp before b2731 contains an out-of-bounds read vulnerability in llama_model_loader that allows attackers to read memory outside the mapped model file via crafted GGUF tensor offsets. Attackers can supply a crafted or truncated GGUF model file with arbitrary 64-bit tensor offsets to crash the process or load adjacent process memory as model weights.
Metrics
CVSS 3.1 6.1/10
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:HWeakness enumeration
Affected software
Source: the vendor's advisory (CNA). NVD analysis pending.
| Vendor | Product | Versions |
|---|---|---|
| ggml-org | llama.cpp | >= b1046, < b2731 |
References
- https://github.com/ggml-org/llama.cpp/blob/b2730/llama.cpp#L3002-L3004
- https://github.com/ggml-org/llama.cpp/commit/0ead1f1072fdc70720f92e008a294dc74a826b1d
- https://github.com/ggml-org/llama.cpp/issues/6774
- https://github.com/ggml-org/llama.cpp/pull/6885
- https://github.com/ggml-org/llama.cpp/releases/tag/b2731
- https://www.vulncheck.com/advisories/llama-cpp-before-b2731-out-of-bounds-read-via-gguf-tensor-offsets
Timeline
- Published
- Oct 27, 2026
- Last modified
- Nov 2, 2026
- NVD status
- Awaiting Analysis