{"id":"CVE-2026-109227","published":"2026-10-12T14:17:11.923","lastModified":"2026-10-12T14:17:12.040","description":"SGLang through 0.5.21 contains an unbounded memory allocation vulnerability in preprocess_video in qwen_vl.py that decodes frames at full source resolution before applying max_pixels limits. Remote attackers can submit high-resolution videos to /v1/chat/completions on Qwen-VL deployments, and concurrent requests exhaust worker memory causing denial of service.","cvssScore":7.5,"cvssSeverity":"HIGH","cvssVector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H","cwes":["CWE-770"],"vendors":[],"products":[],"references":[{"url":"https://github.com/sgl-project/sglang","tags":[]},{"url":"https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/multimodal/processors/qwen_vl.py#L267-L306","tags":[]},{"url":"https://github.com/sgl-project/sglang/issues/43710","tags":[]},{"url":"https://www.vulncheck.com/advisories/sglang-through-0.5.21-memory-exhaustion-dos-via-qwen-vl-video-preprocessing","tags":[]}],"exploitRefs":[{"url":"https://github.com/sgl-project/sglang","tags":[]},{"url":"https://github.com/sgl-project/sglang/blob/v0.5.21/python/sglang/srt/multimodal/processors/qwen_vl.py#L267-L306","tags":[]},{"url":"https://github.com/sgl-project/sglang/issues/43710","tags":[]}],"hasPoc":true,"ai":null}