llama.cpp: Llama.cpp through commit 97f06e9, when started with the --reranking flag, allows remote…
llama.cpp
More for llama.cpp RSS feed for 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
- Record
- CVE-2026-52132
- Affects
- llama.cpp
- Known exploited
- Not listed in the CISA Known Exploited Vulnerabilities catalog
- Published
- Record first published by the source
- September 1, 2026
- Source last modified
- September 4, 2026
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.
Description quoted from the NVD record.
Who is affected
llama.cpp, made by ggml, is a program that runs AI language models and answers requests from other software. This flaw affects llama.cpp up to and including the code change labeled 97f06e9, but only when it is started with its reranking feature turned on (a feature that sorts search results by how relevant they are). Someone who can reach the server over a network could send a specially crafted web request that makes the reranking part of the program fail, so it stops answering requests properly. This flaw is rated High.
What to do
- Ask the person who installed or runs llama.cpp for your organization to check whether you use it and whether it is started with its reranking feature turned on. Only those setups are affected.
- No fixed version is named for this flaw yet, so check the vendor advisory linked on this page for a fixed release. That person should install the fixed release as soon as it is available.
- If an IT provider or vendor runs llama.cpp for your organization, send them this advisory and ask when the update will be done.
Analysis written from the source record as of September 4, 2026 and reviewed before publishing.