CVE-2026-90554

Source
https://cve.org/CVERecord?id=CVE-2026-90554
Import Source
https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-90554.json
JSON Data
https://api.osv.dev/v1/vulns/CVE-2026-90554
Aliases
  • GHSA-936p-m5pv-vvjf
Published
2026-09-12T12:08:57Z
Modified
2026-09-13T03:47:24Z
Severity
  • 6.9 (Medium) CVSS_V4 - CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N CVSS Calculator
Summary
vLLM before 0.28.0 Denial of Service via audio extraction
Details

vLLM versions >=0.10.2 and <0.28.0 do not apply any audio decode-size or duration limit when extracting audio from video input for NanoNemotronVL models. In nano_nemotron_vl.py, _extract_audio_from_videos calls load_audio_pyav(BytesIO(video_bytes)) without the max_duration_s or max_decode_bytes parameters, so neither VLLM_MAX_AUDIO_DECODE_DURATION_S nor VLLM_MAX_AUDIO_DECODE_BYTES is enforced (unlike the direct audio upload path in AudioMediaIO). When a NanoNemotronVL model is served with use_audio_in_video=True, an attacker who supplies a small, highly compressed video as multimodal input can force the server to allocate gigabytes of memory during audio decoding, resulting in a denial of service. Fixed in vLLM 0.28.0.

Database specific
{
    "cna_assigner": "VulnCheck",
    "cwe_ids": [
        "CWE-400"
    ],
    "osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/90xxx/CVE-2026-90554.json"
}
References

Affected packages

Git / github.com/vllm-project/vllm

Affected ranges

Type
GIT
Repo
https://github.com/vllm-project/vllm
Events
Database specific
Show details
{
    "extracted_events": [
        {
            "introduced": "0.10.2"
        },
        {
            "fixed": "0.28.0"
        }
    ],
    "source": "AFFECTED_FIELD"
}

Database specific

source
"https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-90554.json"