CVE-2026-109227
Deferred Deferred - Pending Action

Memory Exhaustion in SGLang via High-Resolution Video Processing

Publication date: 2026-10-12

Last updated on: 2026-10-12

Assigner: VulnCheck

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.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-10-12
Last Modified
2026-10-12
Generated
2026-10-12
AI Q&A
2026-10-12
EPSS Evaluated
N/A
NVD
CVE-2026-109227
EUVD
EUVD-2026-97433

Affected Vendors & Products

Showing 1 associated CPE
Vendor Product Version / Range
sgl-project sglang 0

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-770 The product allocates a reusable resource or group of resources on behalf of an actor without imposing any intended restrictions on the size or number of resources that can be allocated.

Attack-Flow Graph

AI Quick Actions

Instant insights powered by AI
Executive Summary

SGLang through version 0.5.21 has an unbounded memory allocation issue in the preprocess_video function in qwen_vl.py. It decodes video frames at their full resolution before applying any size limits. Attackers can exploit this by sending high-resolution videos to the /v1/chat/completions endpoint on Qwen-VL deployments, causing memory exhaustion and denial of service.

Detection Guidance

Monitor for high memory usage on Qwen-VL deployments, especially during video processing. Check for repeated requests to /v1/chat/completions with large video files. Use system monitoring tools like top, htop, or ps to track memory consumption by SGLang processes.

Impact Analysis

If you run a Qwen-VL deployment using SGLang 0.5.21 or earlier, attackers could send large videos to your system, consuming all available memory. This may crash your service, slow down other applications, or make your system unresponsive, leading to downtime and potential data loss.

Compliance Impact

This vulnerability could lead to service disruptions or data unavailability, which may violate availability requirements in GDPR and HIPAA. Downtime could result in unauthorized access or loss of personal health data, potentially breaching compliance obligations.

Mitigation Strategies

Upgrade SGLang to a version beyond 0.5.21 where the issue is fixed. Implement rate limiting on /v1/chat/completions endpoints to prevent excessive requests. Set memory limits for SGLang processes to prevent denial of service. Validate and limit video resolution before processing.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-109227. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart