CVE-2026-43627
llama.cpp b1283–b9058 Integer Overflow in llama_batch_init() Function
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
Attack profile
How easy it is to exploit
- Attack vector
- Local
- Complexity
- Low
- Privileges needed
- None
- User interaction
- Passive
What an attacker can affect
- Confidentiality
- High
- Integrity
- High
- Availability
- High
Exploit likelihood
What each source reports
CVE List Score used
llama.cpp b1283–b9058 Integer Overflow in llama_batch_init() Function
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
CVE-2026-43627- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N - CVSS 3.1
- 7.8 High
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H - Weakness
- CWE-190, CWE-680
- Updated
- 2026-11-02
EUVDEUVD-2026-54281
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
EUVD-2026-54281- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N - Updated
- 2026-11-02
GitHub AdvisoriesGHSA-6gxr-4p8w-gxc9
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the...
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
GHSA-6gxr-4p8w-gxc9- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - CVSS 3.1
- 7.8 High
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H - Weakness
- CWE-190
- Updated
- 2026-11-02
NVD
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
CVE-2026-43627- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - CVSS 3.1
- 7.8 High
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H - Weakness
- CWE-190, CWE-680
- Updated
- 2026-11-02
OSV
llama.cpp b1283–b9058 Integer Overflow in llama_batch_init() Function
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
CVE-2026-43627- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N - Weakness
- CWE-190, CWE-680
- Updated
- 2026-11-02
OSVDEBIAN-CVE-2026-43627
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
DEBIAN-CVE-2026-43627- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X - Updated
- 2026-11-02
OSVUBUNTU-CVE-2026-43627
llama.cpp builds b1283 through b9058 contain an integer overflow vulnerability in the llama_batch_init() function where unchecked multiplications in malloc() calls can wrap past INT32_MAX when computing allocation sizes. Attackers can pass specially crafted parameters to trigger integer overflow, causing heap corruption and potentially achieving arbitrary code execution through subsequent batch operations that write past allocated buffer boundaries.
- Record
UBUNTU-CVE-2026-43627- CVSS 4.0
- 8.5 High
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N - CVSS 3.1
- 7.8 High
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H - Updated
- 2026-11-02
Affected products
Affected packages
| Package | Affected | Fixed |
|---|---|---|
llama.cppDebian:14 |
All versions, plus listed versions 10108+dfsg-110108+dfsg1-110108+dfsg1-210192+dfsg-110192+dfsg-2 Show 45 moreShow less10192+dfsg-2~exp110271+dfsg-110344+dfsg-110438+dfsg-11:0.1.2+dfsg-1~exp11:0.2.0+dfsg1-11:0.2.0+dfsg1-1~exp11:0.4.0+dfsg-11:0.4.1+dfsg-11:0.5.0+dfsg-11:0.5.0+dfsg-21:0.6.0+dfsg-11:0.6.0+dfsg-21:0.7.0+dfsg-11:0.7.0+dfsg-25151+dfsg-1~exp25151+dfsg-1~exp35318+dfsg-15318+dfsg-25713+dfsg-15760+dfsg-15760+dfsg-25760+dfsg-35760+dfsg-45882+dfsg-15882+dfsg-25882+dfsg-35882+dfsg-3~exp15882+dfsg-3~exp25882+dfsg-3~exp35882+dfsg-46641+dfsg-16641+dfsg-26641+dfsg-37593+dfsg-17593+dfsg-27593+dfsg-37965+dfsg-18064+dfsg-18064+dfsg-28461+dfsg-18611+dfsg-18681+dfsg-18870+dfsg-18941+dfsg-1 |
None known |
llama.cppUbuntu:26.10 |
All versions, plus listed versions 8681+dfsg-19721+dfsg-1ubuntu110438+dfsg-110438+dfsg-1ubuntu1 |
None known |
llama.cppUbuntu:26.04:LTS |
All versions, plus listed versions 5882+dfsg-28064+dfsg-1ubuntu18681+dfsg-1 |
None known |
History
| Recorded | Changed | Sources |
|---|---|---|
| first indexed | CVE List, EUVD, GitHub Advisories, NVD, OSV |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
CVE-2026-43627 is a high vulnerability (CVSS 8.5) in Ggml Llama.cpp, published August 6, 2026. No source lists public exploit code. EPSS puts the chance of exploitation in the next 30 days at 0.20%, the 10th percentile. No source names a fixed version.
References
github.com/Vladimir-tokarev-cyera/llama-cpp-security-patches
vulncheck.com/advisories/llama-cpp-b1283-b9058-integer-overflow-in-llama-batch-init-function
nvd.nist.gov/vuln/detail/CVE-2026-43627
security-tracker.debian.org/tracker/CVE-2026-43627
ubuntu.com/security/CVE-2026-43627
cve.org/CVERecord?id=CVE-2026-43627
Left out here: 13 versions. The JSON record has them.