llama.cpp: Llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in…
llama.cpp
More for llama.cpp RSS feed for llama.cpp
- Severity
- High, CVSS 8.1
- CVSS vector
- CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H
- Record
- CVE-2026-43632
- Affects
- llama.cpp
- Known exploited
- Not listed in the CISA Known Exploited Vulnerabilities catalog
- Published
- Record first published by the source
- August 6, 2026
- Source last modified
- October 23, 2026
Description
llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in llama-server affecting six tokenization endpoints (/tokenize, /detokenize, /infill, /apply-template, /rerank, and /anthropic/count_tokens) that bypass the task queue and access ctx_server.vocab directly on HTTP worker threads. Attackers can exploit a time-of-check-time-of-use race condition where the main thread destroys and frees vocab after the synchronization lock is released but before the handler finishes using it, causing a crash or potential code execution when --sleep-idle-seconds is configured.
Description quoted from the NVD record.
Who is affected
This flaw affects llama.cpp, made by ggml, a program that runs AI language models and answers requests from other software. It affects builds b7492 through b9060, the latest build. The problem is in the built-in server part of the product, in six features that work with text, such as counting it or splitting it into pieces. When the server is set to go to sleep after a period of inactivity, an attacker could make it crash or possibly run their own code on it. This flaw is rated High.
What to do
- The person who installed or runs llama.cpp in your organization should handle this. If an IT provider or vendor manages it for you, they should handle it.
- Find out which build you use. Builds b7492 through b9060 are affected.
- No fixed version is named for this flaw yet. Check the vendor advisory linked on this page for a fixed release, and install it as soon as it is available.
- This flaw only applies when the server is set to sleep after a period of inactivity, so find out whether your setup uses that option.
- If an IT provider or vendor runs llama.cpp for you, send them this advisory and ask when the update will be done.
Analysis written from the source record as of October 23, 2026 and reviewed before publishing.