CVE-2026-115913

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.

• CWE-770

• https://github.com/vllm-project/vllm/commit/3d52578e13454de0bb8352cac988347cdcc89928
• https://github.com/vllm-project/vllm/pull/60614
• https://github.com/vllm-project/vllm/releases/tag/v0.32.0
• https://github.com/vllm-project/vllm/security/advisories/GHSA-8397-8qf6-f2gm

Published

Oct 26, 2026

Last Modified

Oct 26, 2026

Status

Awaiting Analysis

• How Strix found a critical auth bypass in etcdStrix autonomously discovered a critical authentication bypass in etcd, later designated CVE-2026-33413.
• Autonomous PentestingAI agents that find and validate exploitable vulnerabilities like this one across your applications.
• PR ReviewsPentest every pull request so vulnerable code is caught before it ships to production.
• AI Penetration TestingHow AI-driven penetration testing continuously covers your attack surface.

• CVE-2026-115908zip is a zip library for Rust that supports reading and writ…7.4
• CVE-2026-115909zip is a zip library for Rust that supports reading and writ…7.5
• CVE-2026-11591The Widgets for Google Reviews plugin for WordPress is vulne…4.4
• CVE-2026-115910Python Social Auth is a social authentication and registrati…6.8
• CVE-2026-115911vLLM is an inference and serving engine for large language m…6.5
• CVE-2026-115912vLLM is an inference and serving engine for large language m…7.5
• CVE-2026-115914vLLM is an inference and serving engine for large language m…6.5
• CVE-2026-115915vLLM is an inference and serving engine for large language m…8.1
• CVE-2026-115916vLLM is an inference and serving engine for large language m…6.5
• CVE-2026-115917vLLM is an inference and serving engine for large language m…3.4
• CVE-2026-115918vLLM is an inference and serving engine for large language m…6.5
• CVE-2026-115919Open Neural Network Exchange (ONNX) is an open standard for …6.2