---
cve: "CVE-2026-43632"
severity: "CRITICAL"
cvss: 9.2
epss: "0.3%"
vendor: "ggml-org"
kev: false
exploited: false
published: "2026-08-06 22:17:06"
tags: [cve, security, critical]
source: tsecurity.de CVE-Dossier
exported: "2026-09-22T09:38:13+02:00"
---

# CVE-2026-43632

> 9.2 CRITICAL · 🧪 PoC

## Beschreibung

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.

## CVSS-Vektor

```
CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
```

| Metrik | Wert | Bewertung |
|---|---|---|
| AV Angriffsvektor | Netzwerk | bad |
| AC Komplexität | Hoch | good |
| PR Privilegien | Keine | bad |
| UI Interaktion | Keine | bad |

## Referenzen

- <https://github.com/Vladimir-tokarev-cyera/llama-cpp-security-patches>
- <https://www.vulncheck.com/advisories/llama-cpp-b7492-b9060-use-after-free-in-tokenization-endpoints>

---
_Exportiert aus dem [tsecurity.de CVE-Dossier](https://tsecurity.de/cve?cve=CVE-2026-43632) · Datenquellen: EUVD (ENISA), NVD, OSV, CISA KEV, FIRST EPSS, Exploit-DB, BSI BITS_