SOURCE-LINKED INTELLIGENCE
GHSA-6pr9-rp53-2pmc: vLLM: OOM Denial of Service via Audio Decompression Bomb
### Summary vLLM's `/v1/audio/transcriptions` endpoint limits compressed upload size but not decoded PCM output. A 25MB OPUS file expands to ~14.9GB of float32 PCM at decode time. Tested on vLLM v0.19.0. ### Details `SpeechToTextProcessor` rejects uploads over `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB` (default 25MB) based on compressed byte length, but the audio decoder in `audio.py` accumulates all decoded frames into memory with no size limit before returning: ```python # speech_to_text.py L184-189 if len(audio_data) / 1024 ** 2 > self.max_audio_filesize_mb: raise VLLMValidationError(...) y, sr = lo
Read original source ↗ Open in workspace
- recordType
- vulnerability
- status
- active
- evidenceStatus
- reported
- region
- Global
Evidence & attribution
- OSV AI package advisories · 2026-06-17T14:06:22.000Z
- OSV AI package advisories · 2026-07-13T15:46:18.784Z
First collected: 2026-09-20T22:31:48.298Z. This is not the publication date.