CVE-2026-43632
Published: Thu, 06 Aug 2026 22:17
Summary
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,
Details
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.
Are YOU affected by CVE-2026-43632?
5-second check on your actual server. Reads /etc/os-release, uname -r, and dpkg-query; matches against the live USN + Debian Security Tracker feeds; tells you whether CVE-2026-43632 (and any other live CVE) applies. Anonymous, no signup.
curl https://mindsparkstack.com/scan.sh | bash
References
StackPatch runs this match against YOUR installed packages every hour
Free (3 servers) / from $9/mo (14-day free trial) / Solo $9/mo / Pro $29/mo / Team $79/mo. Indie pricing.