HIGHCVSS v3.1 7.8Vulnerability

CVE-2026-133710

llama.cpp b11003 before b12556 contains an integer overflow in the hrm_text loader load_arch_hparams() that allows attackers to bypass the block count check using crafted GGUF metadata. Attackers can supply a GGUF model whose layer factors wrap modulo 2^32, causing load_arch_tensors() to write tensor pointers past model.layers, corrupting heap memory and potentially enabling code execution.

Properties

Severity
HIGH
Nvd fetched at
2026-11-22T12:30:49.958423+00:00
Epss score
0.00313
Cvss severity
HIGH
Cvss score
7.8
Retrieved at
2026-11-22T12:32:26+00:00
Last source
FIRST EPSS
Score
7.8
Cvss version
3.1
Cve id
CVE-2026-133710
Cvss vector
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
First seen by Signal
2026-11-20T15:43:28+00:00
Name
CVE-2026-133710
Vector
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Published at
2026-11-20T13:18:47.340
Last modified
2026-11-20T18:18:51.863
Epss percentile
0.24662

Related Entities (1)

HAS_WEAKNESS (1)

Related research

Explore deeper with Ninja Signal's threat intelligence graph

CVE-2026-133710 (CVSS v3.1 7.8) — Ninja Signal Threat Intelligence | Ninja Signal