CVE-2026-34760

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.

CVSS v3 5.9 MEDIUM

5.9^/10

CVSS v3 : MEDIUM

Vector :

Exploitability : 1.6 / Impact : 4.2

Attack Vector NETWORK

Attack Complexity HIGH

Privileges Required LOW

User Interaction NONE

Confidentiality Impact NONE

Integrity Impact HIGH

Availability Impact LOW

Scope UNCHANGED

References

Link	Resource
https://github.com/vllm-project/vllm/commit/c7f98b4d0a63b32ed939e2b6dfaa8a626e9b46c4	Patch
https://github.com/vllm-project/vllm/pull/37058	Issue Tracking
https://github.com/vllm-project/vllm/releases/tag/v0.18.0	Release Notes
https://github.com/vllm-project/vllm/security/advisories/GHSA-6c4r-fmh3-7rh8	Vendor Advisory

Configurations

Configuration 1 (hide)

cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*

History

No history.

Information

Published : 2026-04-02 20:16

Updated : 2026-07-24 21:10

NVD link : CVE-2026-34760

Mitre link : CVE-2026-34760

CVE.ORG link : CVE-2026-34760

JSON object : View

Products Affected

vllm

vllm

CWE

CWE-20

Improper Input Validation

{"id": "CVE-2026-34760", "cveTags": [], "metrics": {"ssvcV203": [{"source": "134c704f-9b21-4f2e-91b3-4a467353bcc0", "ssvcData": {"id": "CVE-2026-34760", "role": "CISA Coordinator", "options": [{"exploitation": "none"}, {"automatable": "no"}, {"technicalImpact": "partial"}], "version": "2.0.3", "timestamp": "2026-04-03T14:42:25.211772Z"}}], "cvssMetricV31": [{"type": "Secondary", "source": "security-advisories@github.com", "cvssData": {"scope": "UNCHANGED", "version": "3.1", "baseScore": 5.9, "attackVector": "NETWORK", "baseSeverity": "MEDIUM", "vectorString": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:H/A:L", "integrityImpact": "HIGH", "userInteraction": "NONE", "attackComplexity": "HIGH", "availabilityImpact": "LOW", "privilegesRequired": "LOW", "confidentialityImpact": "NONE"}, "impactScore": 4.2, "exploitabilityScore": 1.6}, {"type": "Primary", "source": "nvd@nist.gov", "cvssData": {"scope": "UNCHANGED", "version": "3.1", "baseScore": 7.1, "attackVector": "NETWORK", "baseSeverity": "HIGH", "vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:L", "integrityImpact": "HIGH", "userInteraction": "NONE", "attackComplexity": "LOW", "availabilityImpact": "LOW", "privilegesRequired": "LOW", "confidentialityImpact": "NONE"}, "impactScore": 4.2, "exploitabilityScore": 2.8}]}, "affected": [{"source": "security-advisories@github.com", "affectedData": [{"vendor": "vllm-project", "product": "vllm", "versions": [{"status": "affected", "version": ">= 0.5.5, < 0.18.0"}]}]}], "published": "2026-04-02T20:16:25.437", "references": [{"url": "https://github.com/vllm-project/vllm/commit/c7f98b4d0a63b32ed939e2b6dfaa8a626e9b46c4", "tags": ["Patch"], "source": "security-advisories@github.com"}, {"url": "https://github.com/vllm-project/vllm/pull/37058", "tags": ["Issue Tracking"], "source": "security-advisories@github.com"}, {"url": "https://github.com/vllm-project/vllm/releases/tag/v0.18.0", "tags": ["Release Notes"], "source": "security-advisories@github.com"}, {"url": "https://github.com/vllm-project/vllm/security/advisories/GHSA-6c4r-fmh3-7rh8", "tags": ["Vendor Advisory"], "source": "security-advisories@github.com"}], "vulnStatus": "Analyzed", "weaknesses": [{"type": "Secondary", "source": "security-advisories@github.com", "description": [{"lang": "en", "value": "CWE-20"}]}], "descriptions": [{"lang": "en", "value": "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."}, {"lang": "es", "value": "vLLM es un motor de inferencia y servicio para modelos de lenguaje grandes (LLMs). Desde la versi\u00f3n 0.5.5 hasta antes de la versi\u00f3n 0.18.0, Librosa por defecto utiliza numpy.mean para el downmixing mono (to_mono), mientras que el est\u00e1ndar internacional ITU-R BS.775-4 especifica un algoritmo de downmixing ponderado. Esta discrepancia resulta en inconsistencia entre el audio escuchado por humanos (p. ej., a trav\u00e9s de auriculares/altavoces normales) y el audio procesado por modelos de IA (que infra a trav\u00e9s de Librosa, como vllm, transformer). Este problema ha sido parcheado en la versi\u00f3n 0.18.0."}], "lastModified": "2026-07-24T21:10:00.143", "configurations": [{"nodes": [{"negate": false, "cpeMatch": [{"criteria": "cpe:2.3:a:vllm:vllm:*:*:*:*:*:*:*:*", "vulnerable": true, "matchCriteriaId": "B8A23C5E-0560-4C39-AF88-AA055348DC8B", "versionEndExcluding": "0.18.0", "versionStartIncluding": "0.5.5"}], "operator": "OR"}]}], "sourceIdentifier": "security-advisories@github.com"}