aidc-ai/pixelle-video

Pixelle-Video is an AI-powered short video creation engine that can produce a complete video from a single topic or theme, no video editing experience needed. You type in a subject, and the system automatically writes a script, generates matching images or video clips using AI, synthesizes a voiceover narrator, adds background music, and assembles everything into a finished video. The tool is built in Python and works through a web-based interface. On Windows, a one-click package is available that requires no additional setup. On other systems, you install a few tools and launch the web interface through a browser. The interface has three panels: content input on the left, settings in the middle, and a preview or output area on the right. It supports multiple AI models for script writing, image generation via ComfyUI (a local image generation tool) or a cloud service called RunningHub, and text-to-speech (TTS) voice synthesis through several audio engines. You can choose between vertical and horizontal video formats, swap out visual templates, and supply your own photos or footage as custom material. The project is actively maintained, with features like digital avatar lip-sync, motion transfer, and multi-language voice support added in recent updates. It is aimed at content creators who want to automate short-form video production without touching traditional editing software.