wtf is openwebui-gemini-omni-video?

OpenWebUI Gemini Omni Video Generator is a tool that adds video generation to OpenWebUI, a chat interface for AI models. It uses Google's Gemini Omni Flash model to turn a written prompt into a realistic video, then downloads that video and shows it as a playable video right inside your chat conversation, with no need to leave the app to view it. Beyond making videos from text, the tool can also start from an image you upload or link to, using it as a reference for the generated video. It can also edit an existing video: you attach a short clip, or provide a link to a longer one stored in Google Cloud Storage, and describe the change you want, such as replacing the background or changing the style. Setting this up takes more than a normal plug in install, because it connects to a Google Cloud Project rather than using a simple API key. Depending on how OpenWebUI itself was installed (plain Python, Docker, or Kubernetes), the setup steps differ: a Python install just needs you to log into your Google account through the gcloud command line tool, while Docker and Kubernetes installs require creating a Google Cloud service account, downloading its key file, and pointing your container or cluster at that file through an environment variable or secret. Once authenticated, you install the tool's Python file into OpenWebUI, enter your Google Cloud project ID and region, and turn the tool on for the AI model you want to use it with. The README also includes a detailed prompting guide for getting good results from the video model: how to ask for a single continuous shot instead of multiple cuts, how to keep edits simple and specific so the rest of the video stays unchanged, how to request particular background music, how to time events to happen at certain seconds, how to make readable text appear on screen, and how to use multiple reference images together, for example one as the starting frame and another as a style reference. The project does not state a license in its README.