CVE-2026-55646
MediumCVSS 6.5Exploitation Probability (EPSS)
Low risk21th percentile - higher than 21% of all known CVEs
Summary
A vulnerability in vLLM versions 0.22.0 to 0.23.0 allows an attacker to upload oversized audio files via the /v1/audio/transcriptions or /v1/audio/translations API. The file is read into memory before the size limit is checked, potentially causing memory exhaustion or process termination.
Risk Assessment
The organization is exposed to a DoS (Denial of Service) attack where a large audio file can exhaust server memory and crash vLLM, impacting service availability.
Recommendation
Upgrade vLLM to version 0.24.0 or later immediately, which fixes the issue by checking the file size limit before reading the file into memory.
Original NVD description (English source)
vLLM is an inference and serving engine for large language models. From 0.22.0 to 0.23.0, the /v1/audio/transcriptions and /v1/audio/translations routes call request.file.read() to fully materialize an uploaded audio file into memory before vLLM checks the documented VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed upload size limit (default 25 MB) later in the speech-to-text preprocessing step, so an API caller who can reach those routes can submit an oversized multipart upload and cause vLLM to allocate memory proportional to the uploaded file size before the request is rejected as too large, creating memory pressure or terminating the process depending on deployment resource limits. This issue is fixed in version 0.24.0.