CVE-2025-52566
High
Description
llama.cpp is an inference of several LLM models in C/C++. Prior to version b5721, there is a signed vs. unsigned integer overflow in llama.cpp's tokenizer implementation (llama_vocab::tokenize) (src/llama-vocab.cpp:3036) resulting in unintended behavior in tokens copying size comparison. Allowing heap-overflowing llama.cpp inferencing engine with carefully manipulated text input during tokenization process. This issue has been patched in version b5721.
SeverityHigh
CVSS8.6
Exploit probability—
ProductGgml Llama.cpp
PublishedJun 24, 2025
First seenJun 24, 2025
Weakness typeCWE-119Memory buffer bounds errorCWE-195CWE-195
How to fix
Update
- ggml llama.cppb5721 or later
Affected products
| Product | Affected versions | Fixed in |
|---|---|---|
| ggml llama.cpp | v < b5721 | b5721 |
Check your version
Choose the product and enter the version you run to see whether it falls in the affected range.
In the Intelligence 360 plan
Attack Surface Management
See the assets tied to your domains, the technologies they run and known vulnerabilities in one list.