CVE-2026-126396 · AI agents & MCP
NO FIX YET · SEE WHAT TO DO
vLLM's MiniMax M3 VL video processor accepts a request-level min_pixels value without validation, so a client can enlarge every decoded frame. This drives excessive CPU and host-memory use that can disrupt workers or the service.
first public Oct 27, 2026 · 1 source
Check it with your coding agent
It reads this page and tells you if your code or app is affected.
Teams running vLLM from 0.24.0 before 0.31.0 that serve MiniMax M3 VL video inputs to API clients. Exposure is greatest when the video pixel budget (max_long_side_pixel) is left at its default of unset and clients can pass mm_processor_kwargs. Deployments on other models or on versions outside the range can skip it.
Not affected if no-minimax-m3-vl, no-video-input, outside-version-range
Who cares ai-inference, minimax-m3-vl, self-hosted, vllm
pip show vllm and compare it to the fixed range.mm_processor_kwargs from API clients.max_long_side_pixel is set in the video processor configuration; it defaults to None.0.31.0 or later.vLLM from 0.24.0 before 0.31.0 contains a resource exhaustion vulnerability in the MiniMax M3 VL video processor, which accepts request-level min_pixels without range or min_pixels <= max_pixels validation. Its video-wide pixel budget is enforced only when max_long_side_pixel is set, and that value defaults to None. API clients can supply a large min_pixels via mm_processor_kwargs to enlarge every decoded frame, causing excessive CPU and host-memory consumption that may disrupt workers or the service.
This page as .md · .json · CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X
Are you affected?
Ask from here. InfraStack checks the affected versions against your stack and answers in your inbox.
InfraStack checks the affected versions against your stack and answers in Slack, with the evidence and the fix.
InfraStack checks it against your stack and answers on your report, with the evidence and the fix.
NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.