CVE-2026-115913
mediumpriority · track · 0EPSS 0.003 · 23 %ileDEUSEPSS · scored 2026-10-28
vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the Phi-4 multimodal audio encoder in phi4mm_audio.py and phi4mm_utils.py uses eager attention by default, allowing a remote client to submit one legal long audio clip that causes MultiHeadedAttention.forward_attention to materialize and retain full attention matrices in self.attn. The scaled dot-product attention branch does not materialize these matrices. The resulting quadratic GPU-memory use can cause an out-of-memory condition and terminate EngineCore for all users. This issue is fixed in version 0.32.0.
cvss
Technical severity, 0–10. One row per scoring source.
base score 0–10 · higher = more severe · hover a vector to decode
weaknesses
The CWE weakness classes this flaw belongs to.
affected products (2)
Products and version ranges known to be affected.
advisories (1)
Vendor and ecosystem advisories describing this.
- WID-SEC-W-2026-4051vllm: Mehrere SchwachstellenDeutsch2026-10-19csaf · cert_bundDE
Ein Angreifer kann mehrere Schwachstellen in vllm ausnutzen, um beliebigen Programmcode auszuführen, um einen Denial of Service Angriff durchzuführen, um Informationen offenzulegen und um Sicherheitsvorkehrungen zu umgehen.
sources
Which sources assert this record, and when.
references (4) nvd
External citations — patches, advisories, analysis.
- https://github.com/vllm-project/vllm/commit/3d52578e13454de0bb8352cac988347cdcc89928Patch
- https://github.com/vllm-project/vllm/pull/60614Patch
- https://github.com/vllm-project/vllm/releases/tag/v0.32.0Release Notes
- https://github.com/vllm-project/vllm/security/advisories/GHSA-8397-8qf6-f2gmExploitPatchVendor Advisory
Attackers weaponise vulnerabilities like this against internet-exposed services. See what is exposed on your perimeter with a free passive scan.
Free scan