CVE-2026-70639: CWE-476 NULL Pointer Dereference in ggml-org llama.cpp
A null pointer dereference vulnerability exists in llama.cpp version 0.9.0 within the LLaMA-Android JNI wrapper. The bench_1model() function does not validate the model context pointer before dereferencing it. Supplying a malicious or corrupted model file can cause a SIGSEGV crash, terminating the Android application and causing denial of service.
AI Analysis
Technical Summary
llama.cpp versions 0.9.0 contain a null pointer dereference vulnerability (CWE-476) in the LLaMA-Android JNI wrapper's bench_1model() function. This function fails to check if the model context pointer is null before dereferencing it. An attacker can exploit this by providing a malicious, corrupt, or truncated model file that triggers a null context condition, causing the application to crash with a SIGSEGV signal. This results in denial of service by terminating the Android application process.
Potential Impact
Exploitation of this vulnerability leads to a denial of service condition by crashing the Android application using the vulnerable llama.cpp library. There is no indication of privilege escalation, data disclosure, or code execution. The impact is limited to application termination.
Mitigation Recommendations
No official patch or remediation is currently available for this vulnerability. Users should monitor the vendor's advisories for updates. Until a fix is released, avoid using untrusted or malformed model files with the affected version to reduce risk.
CVE-2026-70639: CWE-476 NULL Pointer Dereference in ggml-org llama.cpp
Description
A null pointer dereference vulnerability exists in llama.cpp version 0.9.0 within the LLaMA-Android JNI wrapper. The bench_1model() function does not validate the model context pointer before dereferencing it. Supplying a malicious or corrupted model file can cause a SIGSEGV crash, terminating the Android application and causing denial of service.
CVSS v4.0
Score 6.8medium
Affected software
pkg:github/ggml-org/llama.cppRun on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
llama.cpp versions 0.9.0 contain a null pointer dereference vulnerability (CWE-476) in the LLaMA-Android JNI wrapper's bench_1model() function. This function fails to check if the model context pointer is null before dereferencing it. An attacker can exploit this by providing a malicious, corrupt, or truncated model file that triggers a null context condition, causing the application to crash with a SIGSEGV signal. This results in denial of service by terminating the Android application process.
Potential Impact
Exploitation of this vulnerability leads to a denial of service condition by crashing the Android application using the vulnerable llama.cpp library. There is no indication of privilege escalation, data disclosure, or code execution. The impact is limited to application termination.
Mitigation Recommendations
No official patch or remediation is currently available for this vulnerability. Users should monitor the vendor's advisories for updates. Until a fix is released, avoid using untrusted or malformed model files with the affected version to reduce risk.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- VulnCheck
- Date Reserved
- 2026-08-04T20:17:18.298Z
- Cvss Version
- 4.0
- State
- PUBLISHED
- Remediation Level
- null
Threat ID: 6a75070dbf8831d5395f5c76
Added to database: 08/06/2026, 22:13:33 UTC
Last enriched: 08/06/2026, 22:42:30 UTC
Last updated: 08/07/2026, 02:25:30 UTC
Views: 4
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.