CVE 8.8 HIGH

llama.cpp has Out-of-bounds Write in llama-server_CVE-2026-21869

January 7, 2026 invoker category_cve

8.8 / 10

HIGH

CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Description

llama.cpp is an inference of several LLM models in C/C++. In commits 55d4206c8 and prior, the n_discard parameter is parsed directly from JSON input in the llama.cpp server's completion endpoints without validation to ensure it's non-negative. When a negative value is supplied and the context fills up, llama_memory_seq_rm/add receives a reversed range and negative offset, causing out-of-bounds memory writes in the token evaluation loop. This deterministic memory corruption can crash the process or enable remote code execution (RCE). There is no fix at the time of publication.

AI Analysis

Out-of-bounds memory write due to unvalidated input in llama.cpp server's completion endpoints, potentially allowing remote code execution (RCE).

Basic Information

ID CVE-2026-21869

Source GitHub_M

Published Jan 7, 2026 at 23:37

Affected Product

Vendor ggml-org

Product llama.cpp

Version <= 55d4206c8

Affected Versions ggml-org llama.cpp <= 55d4206c8

CWE Classification

CWE-787

AI Assessment

AI Score 8.8 / 10

AI Severity High

Vendor ggml-org

Product llama.cpp

Version <= 55d4206c8

References

github.com /ggml-org/llama.cpp/security/advisories/GHSA-8947-pfff-2f3c

{
    "lastseen": "",
    "description": "llama.cpp is an inference of several LLM models in C/C++. In commits 55d4206c8 and prior, the n_discard parameter is parsed directly from JSON input in the llama.cpp server's completion endpoints without validation to ensure it's non-negative. When a negative value is supplied and the context fills up, llama_memory_seq_rm/add receives a reversed range and negative offset, causing out-of-bounds memory writes in the token evaluation loop. This deterministic memory corruption can crash the process or enable remote code execution (RCE). There is no fix at the time of publication.",
    "published": "2026-01-07T23:37:59.886Z",
    "modified": "2026-01-07T23:37:59.886Z",
    "type": "cve",
    "title": "llama.cpp has Out-of-bounds Write in llama-server",
    "source": "GitHub_M",
    "references": "https://github.com/ggml-org/llama.cpp/security/advisories/GHSA-8947-pfff-2f3c",
    "id": "CVE-2026-21869",
    "bulletinFamily": "",
    "cwe": [
        "CWE-787"
    ],
    "cvelist": null,
    "sourceData": "ggml-org llama.cpp <= 55d4206c8",
    "sourceHref": "",
    "cvss": {
        "score": 8.8,
        "severity": "HIGH",
        "vector": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
        "version": "3.1"
    },
    "cvss2": [],
    "cvss3": {
        "version": "",
        "vectorString": "",
        "baseScore": 0,
        "baseSeverity": "",
        "attackVector": "",
        "attackComplexity": "",
        "privilegesRequired": "",
        "userInteraction": "",
        "scope": "",
        "confidentialityImpact": "",
        "integrityImpact": "",
        "availabilityImpact": "",
        "cvssV3": {
            "version": "",
            "vectorString": "",
            "baseScore": 0,
            "baseSeverity": "",
            "attackVector": "",
            "attackComplexity": "",
            "privilegesRequired": "",
            "userInteraction": "",
            "scope": "",
            "confidentialityImpact": "",
            "integrityImpact": "",
            "availabilityImpact": ""
        }
    },
    "href": "",
    "category_name": "CVE",
    "post_link": "",
    "product": "llama.cpp",
    "version": "<= 55d4206c8",
    "vendor": "ggml-org",
    "ai_description": "Out-of-bounds memory write due to unvalidated input in llama.cpp server's completion endpoints, potentially allowing remote code execution (RCE).",
    "ai_severity": "High",
    "ai_vendor": "ggml-org",
    "ai_product": "llama.cpp",
    "ai_version": "<= 55d4206c8",
    "ai_score": 8.8
}

💭 Join the Security Discussion ❌ Cancel Reply