CVE-2026-57173: vLLM is an inference and serving engine for large language models. Prior to 0.24.0, the input_audio …
- Occurred
- Sep 16, 2026
- Published
- Sep 16, 2026
- Source
- NVD CVE
Summary
vLLM is an inference and serving engine for large language models. Prior to 0.24.0, the input_audio handling path for /v1/chat/completions calls AudioMediaIO.load_bytes or AudioMediaIO.load_file without passing VLLM_MAX_AUDIO_DECODE_DURATION_S to the shared audio decoder. An unauthenticated client can therefore submit a small compressed audio input that expands into a very large float32 PCM allocation, bypassing the duration guard already used by /v1/audio/transcriptions and causing an out-of-memory worker crash. Inline data URLs reach this path without being bounded by VLLM_AUDIO_FETCH_TIMEOUT. The issue affects deployments serving an audio-capable model, and authentication changes only the deployment-specific reachability. This issue is fixed in version 0.24.0.
Tags
- CVE
- CVE-2026-57173
- cvss:MEDIUM
Classification
Rationale: CVE describes a denial-of-service vulnerability in vLLM audio handling causing OOM crashes in LLM inference servers.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8