# CVE-2026-90554

## Summary

- **CVE ID:** CVE-2026-90554
- **Severity:** MEDIUM
- **CVSS Score:** 6.9 (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)
- **CWE:** CWE-400
- **Published:** Sep 12, 2026
- **Last Modified:** Sep 14, 2026

## Description

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.

## Affected Products

- vllm-project — vLLM (0.10.2)
- vllm-project — vLLM (0.28.0)

## References

- [CNA](https://github.com/vllm-project/vllm/security/advisories/GHSA-936p-m5pv-vvjf)
- [CNA](https://www.vulncheck.com/advisories/vllm-before-0.28.0-denial-of-service-via-audio-extraction)

## Exploitation Prediction (EPSS)

- **EPSS Score:** 0.11%
- **EPSS Percentile:** 1.6

---
_Exported from OnDuty AI Vulnerability Intelligence on 2026-09-18._