CVE-2026-18581: Reachable assertion vulnerability in llama.cpp parser
llama.cpp
A vulnerability exists in the Jinja Minja Template Parser component of llama.cpp affecting file parsing functionality. Crafted input can trigger a reachable assertion requiring local access.
disclosed CVENo confirmed exploitation, so this record is tracked as a disclosed vulnerability and not counted as an incident. It moves into the counts if CISA or the vendor confirms exploitation. Why
- CVSS
- 3.3
- EPSS
- <0.01 · top 95% · 2026-11-13
- fix status
- Fix status unknown
- first report
- 3 Aug 2026
- last update
- 3 Aug 2026
- CVE ids
Timeline · 1
Sources · 1
Advisories and databases
- 2026-08-03CVE-2026-18581: A vulnerability was determined in ggml-org llama.cpp e15efe0.NVD · nvd.nist.gov/vuln/detail/CVE-2026-18581
Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.
Classification
- lane
- Patch now
- AI role
- Attack on AI
- categories
- Infrastructure vulnerability
- severity
- 2 · Disclosed vulnerability
- fix status
- Fix status unknown
- targets
- Open-weight model, Framework / tooling
- owasp llm top 10
- LLM03:2025 Supply Chain
- mitre atlas
- no technique for these categories
- dates
- happened 3 Aug 2026 · first reported 3 Aug 2026
- origin
- NVDVulnerability record from NVD / the GitHub Advisory Database. This product uses the NVD API but is not endorsed or certified by the NVD.
- record
- confidence 0.95 · created 2026-10-18 10:04Z · updated 2026-10-18 10:04Z
Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned
Copy a brief
This record as a short Markdown brief, for a ticket, a chat or a coding agent.
the brief as text
# CVE-2026-18581: Reachable assertion vulnerability in llama.cpp parser https://gcdradar.vercel.app/incidents/cve-2026-18581-reachable-assertion-vulnerability-in-llama-cp · AIR-2026-1239 · closed 2026-08-03 - Lane: Patch now - Severity: 2 (Disclosed vulnerability) - Fix status: Fix status unknown - First reported: 2026-08-03 - AI role: Attack on AI - Products: llama.cpp - CVE ids: CVE-2026-18581 - CVSS: 3.3 ## Summary A vulnerability exists in the Jinja Minja Template Parser component of llama.cpp affecting file parsing functionality. Crafted input can trigger a reachable assertion requiring local access. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2026-08-03: CVE-2026-18581: A vulnerability was determined in ggml-org llama.cpp e15efe0. (https://nvd.nist.gov/vuln/detail/CVE-2026-18581) ## Sources - 2026-08-03 NVD: CVE-2026-18581: A vulnerability was determined in ggml-org llama.cpp e15efe0. https://nvd.nist.gov/vuln/detail/CVE-2026-18581 Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closed6.9llama.cpp CVE-2026-78148: llama.cpp RPC Server Null Pointer Dereference Vulnerability
Patch now · Disclosed · since 24 Aug 2026 · CVE-2026-78148
AI - closed7.3llama.cpp CVE-2026-78147: llama.cpp RPC server vulnerable to insecure deserialization
Patch now · Disclosed · since 23 Aug 2026 · CVE-2026-78147
AI - closed6.9llama.cpp CVE-2026-116467: llama.cpp Mirostat Sampler assertion vulnerability
Patch now · Disclosed · No patch yet · since 26 Oct 2026 · CVE-2026-116467
AI - closed7.5llama.cpp llama.cpp vulnerable to reachable assertion via gguf reader
Patch now · Disclosed · since 1 Sep 2026 · CVE-2026-52131
AI - closed7.5llama.cpp llama.cpp vulnerable to uncontrolled recursion and denial of service
Patch now · Disclosed · since 1 Sep 2026 · CVE-2026-52130
AI - closed7.5llama.cpp llama.cpp vulnerable to denial of service via negative top_n reranking value
Patch now · Disclosed · since 1 Sep 2026 · CVE-2026-52132
AI