CVE•Published 2026-10-27•Modified 2026-10-28•0 articles on news•4 live references•NVD data
CVE-2025-71439
Vulnerability data via NVD (ingested)
CVSS v3.1
5.5
MEDIUM
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H
EPSS percentile
2
Exploit Prediction Scoring System · top 98% of all CVEs
Weaknesses (CWE)
Description
llama.cpp before b5649 contains a heap out-of-bounds read vulnerability in llama_vocab::impl::print_info() that allows attackers to crash model loading via oversized special token IDs. Attackers can supply a crafted GGUF model with a short tokenizer.ggml.tokens array, causing llama-cli or llama-server to read past id_to_token and crash.
Timeline
Published 2026-10-27
Modified 2026-10-28
External references
Search for exposed instances
Shodan + Censys queries derived from NVD's CPE data. The vuln tag catches assets Shodan has explicitly linked to this CVE; the product / banner fingerprints find exposed instances even when the vuln tag was never applied (which is common).
More intel sources (5)
Shodan report
vuln:CVE-2025-71439Country / ASN / product breakdown for the vuln query.
Censys
vulnerabilities.cve_id: CVE-2025-71439Censys host search filtered to this CVE id.
grep.app
CVE-2025-71439Public source-code mentions — fast PoC discovery.
GitHub code
CVE-2025-71439GitHub code search for direct mentions.
Google dork
"CVE-2025-71439" exploit -site:nvd.nist.govWrite-ups and news, NVD excluded.
Known PoCs on GitHub
No public proof-of-concept repositories found for CVE-2025-71439 on GitHub.
We haven't classified any articles referencing CVE-2025-71439 yet. The external references above still apply.