vLLM Vulnerabilities

pipLLM Inference

AI Threat Alert tracks 188 known vulnerabilities in vLLM, 13 rated critical — an AI/ML llm inference in the pip ecosystem. Each CVE includes CVSS severity, EPSS exploit probability, patch status, and CISO-grade analysis.

61
Risk Score
188
Total CVEs
13
Critical
pip
Ecosystem
Nov 18, 2026
Last CVE
26%
Patch Rate
37d
Avg Time to Patch
96,942 stars24,834 forks10,395 issues93 dependentsLast push Nov 15, 2026
View on GitHub

Known Vulnerabilities(188 total, page 5 of 8)

SeverityCVE IDSummaryCVSSPublished
MEDIUMCVE-2026-73556vLLM: ReDoS in structured output stalls engine5.3Aug 13, 2026MEDIUMCVE-2026-73555vLLM: verbose errors leak host paths & usernames5.3Aug 13, 2026HIGHCVE-2026-18621OpenShift AI Pipelines: confused deputy grants node-root7.6Aug 10, 2026HIGHCVE-2023-52355libtiff: OOM DoS in vLLM inference container images7.5Jan 25, 2024HIGHCVE-2026-55574vLLM: unbounded regex compile hangs inference worker (ReDoS)7.5Jul 6, 2026MEDIUMCVE-2026-55514vLLM: malformed prompt embeds crash serving via M-RoPE6.5Jul 6, 2026HIGHCVE-2026-54234vLLM: crafted spec-decoding request crashes GPU worker (DoS)7.5Jul 6, 2026MEDIUMCVE-2026-55646vLLM: audio upload OOMs process before size check6.5Jul 6, 2026HIGHCVE-2026-33845GnuTLS: DTLS integer underflow enables OOB read/DoS7.5Apr 30, 2026HIGHCVE-2026-42009gnutls: DTLS packet-reorder bug DoS hits AI inference servers7.5May 18, 2026HIGHCVE-2026-33846GnuTLS: DTLS fragment heap overflow, DoS on AI inference7.5May 4, 2026HIGHCVE-2026-42010gnutls: NUL-byte username bypasses RSA-PSK auth7.1May 7, 2026HIGHCVE-2026-56209libaom: arbitrary address write in AV1 SVC codec7.1Jun 19, 2026HIGHCVE-2026-56211libaom: AV1 SVC OOB write enables RCE7.1Jun 19, 2026HIGHCVE-2026-56210libaom: AV1 SVC bounds-check miss leaks heap, crashes7.1Jun 19, 2026HIGHCVE-2026-56208libaom: heap overflow in AV1 encoder LAP mode7.6Jun 19, 2026MEDIUMCVE-2026-4878libcap: TOCTOU race in cap_set_file() enables privesc6.7Apr 9, 2026HIGHCVE-2026-4775libtiff: integer overflow causes heap OOB write7.8Mar 24, 2026HIGHCVE-2026-10118Poppler: PDF integer overflow enables heap RCE7.8Jun 1, 2026HIGHCVE-2025-5318libssh: OOB read in SFTP handle leaks memory8.1Jun 24, 2025HIGHCVE-2025-9900libtiff: arbitrary write via crafted TIFF image8.8Sep 23, 2025HIGHCVE-2026-54232vLLM: dependency confusion RCE backdoors container images8.8Jun 22, 2026HIGHCVE-2026-56340vLLM: sparse tensor DoS/memory corruption via embeddings7.5Jun 20, 2026HIGHCVE-2025-71379vLLM: ReDoS via crafted API input causes DoS7.5Jun 20, 2026MEDIUMCVE-2026-12706FFmpeg RASC: UAF in decoder crashes AI inference containers6.5Jun 19, 2026

Showing 101–125 of 188

Frequently asked questions

What is vLLM?

vLLM is an AI/ML llm inference tracked by AI Threat Alert for security vulnerabilities in the pip ecosystem.

How many known vulnerabilities does vLLM have?

vLLM has 188 known CVEs, 13 of them critical, tracked from NVD and GitHub Advisory.

Which ecosystem is vLLM distributed in?

vLLM is distributed via the pip ecosystem and categorized as llm inference.

Where does the vLLM vulnerability data come from?

Vulnerability data is sourced from NVD and GitHub Advisory, enriched with CVSS, EPSS, exploit signals, and patch status for each CVE.

How do I assess the risk of vLLM?

Review each CVE below — every entry shows CVSS severity, EPSS exploit probability, exploitation signals, and whether a patched version is available.

Monitor vLLM in your stack

Get instant alerts when new vulnerabilities affect vLLM. CISO analysis, ATLAS technique mappings, and compliance reports included.

Start Monitoring