CVE-2026-34760
MediumCVSS 5.9Exploitation Probability (EPSS)
Low risk19th percentile - higher than 19% of all known CVEs
Summary
vLLM from version 0.5.5 to before 0.18.0 defaults to using numpy.mean for mono downmixing (to_mono), while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This causes inconsistency between audio heard by humans and audio processed by AI models (e.g., via vllm, transformer). This issue is fixed in version 0.18.0.
Risk Assessment
Inconsistency in audio processing can lead to incorrect results in AI applications, affecting the quality and reliability of generated responses.
Recommendation
Upgrade vLLM to version 0.18.0 or later.
Original NVD description (English source)
vLLM is an inference and serving engine for large language models (LLMs). From version 0.5.5 to before version 0.18.0, Librosa defaults to using numpy.mean for mono downmixing (to_mono), while the international standard ITU-R BS.775-4 specifies a weighted downmixing algorithm. This discrepancy results in inconsistency between audio heard by humans (e.g., through headphones/regular speakers) and audio processed by AI models (Which infra via Librosa, such as vllm, transformer). This issue has been patched in version 0.18.0.