Buffer overflow in Ggml-org Llama.cpp
CVE-2026-70638
llama.cpp builds b1886 through b7445 contain an integer overflow vulnerability in the LLaMA-Android JNI wrapper where the new_1batch() function multiplies sizeof(llama_seq_id) by an attacker-controlled n_seq_max parameter without overflow…
Vulnerability class: Buffer Overflow
CVSS v3 metric
CVSS v3 base score 7.8 (High). Vector: CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H.
Affected products
- Ggml-org Llama.cpp — versions b1886, 0.9.0, 5c0d18881e0e9794c96b2602736b758bac9d9388
Weakness classification (CWE)
Public proof-of-concept exploits
References
- disclosure@vulncheck.com (related)
- disclosure@vulncheck.com (patch)
- disclosure@vulncheck.com (release-notes)
- disclosure@vulncheck.com (third-party-advisory)
Frequently asked questions
- What is CVE-2026-70638?
- CVE-2026-70638 is a high-severity vulnerability in Ggml-org Llama.cpp, classified under Heap-based Buffer Overflow. CVSS score: 7.8/10. Published 2026-08-06.
- How severe is CVE-2026-70638?
- High severity. CVSS v3 base score is 7.8 out of 10.
- Is CVE-2026-70638 known to be exploited?
- 1 public proof-of-concept repositories are indexed. Not currently listed in the CISA KEV catalog.