CVE-2026-116816 | 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.
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 whose negative parsed value writes token ids before the bytes_ buffer during model loading, causing crashes or potential code execution.
- 🔗 github.com/ggml-org/llama.cpp/pull/25969
- 🔗 github.com/ggml-org/llama.cpp/pull/26511
- 🔗 github.com/ggml-org/llama.cpp/commit/cf06ad7dfe79e7b…
- 🔗 github.com/ggml-org/llama.cpp/blob/b10259/src/llama-…
- 🔗 github.com/ggml-org/llama.cpp/releases/tag/b10261
- 🔗 www.vulncheck.com/advisories/llama-cpp-b5899-before-b10261-…
Zero-Day & Vulnerability Intelligence Hub
Echtzeit-Tracking mit EPSS Exploit-Wahrscheinlichkeiten, Angriffsvektor-Decodern und KI-Patch-Anleitungen.
📊 Historien-Charts — Criticals-Trend · Vendors · EPSS-Verteilung
| Tier | CVEs | Anteil |
|---|---|---|
| ≥90 % | 740 | 0,2 % |
| ≥50 % | 3.215 | 0,8 % |
| ≥10 % | 22.028 | 5,8 % |
| <10 % | 354.746 | 93,2 % |
CVE-2026-116816 | 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.
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 whose nega