CVE-2026-126394 · AI agents & MCP
NO FIX YET · SEE WHAT TO DO
vLLM 0.30.0 before 0.31.0 has a video backend that takes the number of frames from the request without an upper limit. A crafted request with a huge frame count makes the server allocate large memory and use CPU until the API server is killed for running out of memory.
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 0.30.0 through 0.31.0 exclusive as an API server that accepts chat completion or tokenize requests from clients they do not fully trust. Deployments that are not serving the openpangu video backend are less exposed, but the sources do not say whether the flaw needs that backend to be the one selected.
Not affected if no-public-api, version-outside-range
Who cares python, self-hosted, vllm
pip show vllm and see whether it is in 0.30.0 up to but not including 0.31.0.0.31.0 or later.vLLM 0.30.0 before 0.31.0 contains a resource exhaustion vulnerability in OpenCVDynamicOpenPanguVideoBackend, which uses request-controlled num_frames without an effective ceiling. Unauthenticated attackers can send /v1/chat/completions or /tokenize requests selecting the openpangu video_backend with a huge num_frames, forcing np.linspace allocations and CPU work that OOM-kill the API server.
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
. Posts quoted belong to their authors.
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.