CVEvulnerability

CVE-2026-57173

6.5 MEDIUM Published 2026-09-16 Modified 2026-10-07 Official page ↗
Status
Analyzed
Published
2026-09-16
Last modified
2026-10-07
Assigner
[email protected]
CVSS 3.1 ([email protected])
6.5 MEDIUM — CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Weaknesses
CWE-770

Description

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.

Affected configurations

• vllm vllm (< 0.24.0)

Related

Weakness · CWE (1)

  • CWE-770 Allocation of Resources Without Limits or Throttling weakness

References