wtf is openwebui-veo-video?

This project is a tool for OpenWebUI, a self-hosted chat interface, that lets you generate videos using Google's Veo 3.1 model without leaving the chat window. Instead of switching to a separate app, you can type a prompt and have a realistic, cinematic video appear directly inside the conversation as a playable video, with a download link underneath. Beyond plain text-to-video, the tool supports uploading an image or short video as a starting point or style reference, and it can edit existing videos too, including ones stored in Google Cloud Storage. Users get a settings panel where they can pick the aspect ratio, video length in seconds, and resolution up to 4K, and the tool can generate up to four videos from one request, stacking the results together in the chat. Because video generation can take a while, especially at higher resolutions, the tool works in the background using polling so the chat interface does not freeze while it waits. Setting it up requires more than an API key. It needs a Google Cloud project with the Gemini Enterprise Agent Platform, formerly known as Vertex AI, enabled, and proper Google Cloud identity credentials rather than a simple key. The README gives separate setup steps depending on whether OpenWebUI is installed through Python, Docker, or Kubernetes, each involving creating and connecting a Google Cloud service account. This tool is aimed at people already running OpenWebUI who want video generation folded into their existing chat setup, and who are comfortable working with Google Cloud accounts and credentials. It assumes some familiarity with self-hosting and cloud infrastructure rather than being a plug and play consumer app.