Description
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 products
| Vendor | Product | Affected | Fixed |
|---|---|---|---|
| ggml-org | llama.cpp | < b5649 | b5649 |
Fixed in Linux distributions
Distributions backport fixes, so the version that fixes this vulnerability differs per release — upgrade to at least the version shown. From the distributions' security trackers and advisories.
Debian
Security tracker ↗| Release | Package | Fixed in | Advisory |
|---|---|---|---|
| Debian 14 | llama.cpp | 5713+dfsg-1 | — |
Ubuntu
Security tracker ↗| Release | Package | Fixed in | Advisory |
|---|---|---|---|
| Ubuntu 26.10 | llama.cpp | No fix yet | — |
| Ubuntu 26.04 LTS | llama.cpp | No fix yet | — |
References
- GitHub Security Advisory (GHSA-g4cc-763q-h9h6) vendor-advisory
- https://github.com/ggml-org/llama.cpp/pull/14145 patch issue-tracking
- https://github.com/ggml-org/llama.cpp/commit/c33fe8b8c4427202706b1434e9fc8ab5752c9cac patch
- https://github.com/ggml-org/llama.cpp/blob/b5648/src/llama-vocab.cpp#L2771 technical-description
- llama.cpp b5649 Release Notes release-notes
- VulnCheck Advisory: llama.cpp before b5649 Heap Out-of-Bounds Read via llama_vocab::impl::print_info third-party-advisory