High severity7.8NVD Advisory· Published Oct 27, 2026
CVE-2026-116816
CVE-2026-116816
Description
llama.cpp b5899 before b10261 contains a heap out-of-bounds write vulnerability in llm_tokenizer_plamo2::build that allows attackers to corrupt heap memory via crafted GGUF models. Attackers can supply BYTE-typed token text like <0x-1> whose negative parsed value writes token ids before the bytes_ buffer during model loading, causing crashes or potential code execution.
Affected products
1Patches
Vulnerability mechanics
References
6- github.com/ggml-org/llama.cpp/blob/b10259/src/llama-vocab.cppnvd
- github.com/ggml-org/llama.cpp/commit/cf06ad7dfe79e7bf2934150d4fefe1ce8db05490nvd
- github.com/ggml-org/llama.cpp/pull/25969nvd
- github.com/ggml-org/llama.cpp/pull/26511nvd
- github.com/ggml-org/llama.cpp/releases/tag/b10261nvd
- www.vulncheck.com/advisories/llama-cpp-b5899-before-b10261-heap-out-of-bounds-write-via-plamo2-tokenizernvd
News mentions
0No linked articles in our index yet.