| Component | llama.cpp |
| Severity | high (CVSS 7.5) |
| CVSS vector | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H |
| Published (NVD) | 2026-09-01 |
| Last modified | 2026-09-04 |
| Affected versions | llama.cpp <= 0.4.0 |
| Fixed in | see the vendor advisory |
| Weakness | CWE-674 |
| CISA KEV | not listed |
NVD description
llama.cpp through commit 97f06e9, when started with the --reranking flag, allows remote attackers to cause a denial of service (std::bad_alloc and HTTP 500) via a negative top_n value in a POST request to /rerank.
What to do
- NVD lists no fixed version yet: follow the vendor advisory below and keep llama.cpp off the public internet until a fix ships.
- Do not expose the component's API or UI to the internet without authentication in front of it.
Sources
- NVD: CVE-2026-52132
- https://blog.ph4nt0m.xyz/ko/cves/cve-2026-52132/
- https://github.com/ggml-org/llama.cpp
Other llama.cpp CVEs
llama.cpp before b11393 contains a use-after-free and double free vulnerability in common_chat_peg_mapper::map that allows unauthenticated r...
A vulnerability was detected in ggml-org llama.cpp up to 0.4.0. This impacts the function rpc_server::deserialize_tensor of the file ggml/sr...
llama.cpp b5693 and before is vulnerable to Uncontrolled Recursion in common/json-schema-to-grammar.cpp, resulting in a denial of service.
llama.cpp b5693 and before has a Reachable Assertion via the gguf_reader::read function.
A vulnerability was determined in ggml-org llama.cpp bec4772f6. This affects the function rpc_server::graph_compute of the file ggml/src/ggm...
A vulnerability was found in ggml-org llama.cpp bec4772f6. The impacted element is the function deserialize_tensor of the file ggml/src/ggml...