CVE-2026-90554: medium-severity vulnerability in vllm-project vLLM
vLLM before 0.28.0 Denial of Service via audio extraction
Published · Updated
E-mail alert when it lands in KEV, gets a public PoC or its score goes up.
13Vexday Risk Score
No sign of exploitation. No public exploitation artifact known so far.
ssvc Trackcvss 6.9epss 0.2%
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.
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
Affected products
vllm-project · vLLMRelated CVEs — vllm-project vLLM
In the same product, most dangerous first.
CVE-2026-103241MEDIUMvllm-project vLLM Gemma4UnifiedParser gemma4.rs denial of serviceEPSS 0.7%CVE-2026-92220MEDIUMvllm-project vLLM MoRIIO Acknowledgement moriio_connector.py MoRIIOWrapper._handle_release_message resource consumptionEPSS 0.7%CVE-2026-90878MEDIUMvllm-project vLLM Jinja Template Rendering completions resource consumptionEPSS 0.5%CVE-2026-90555HIGHvLLM before 0.28.0 Denial of Service via Audio HeaderEPSS 0.5%CVE-2026-105922MEDIUMvllm-project vLLM Penalty utils.py get_token_bin_counts_and_mask denial of serviceEPSS 0.3%CVE-2026-90553HIGHvLLM before 0.28.0 Remote Code Execution via LlavaOnevision2 processorEPSS 0.3%