Pixelle-Video

Why we included this project

Pixelle-Video turns a single topic prompt into a finished short video with no manual editing. You give it a subject and it writes the script, generates matching images or video segments, synthesizes a voiceover, adds background music, and assembles the result into a playable file. The engine is modular rather than tied to one provider: image, video, and TTS steps can run through a local ComfyUI setup, a cloud service like RunningHub, or direct APIs from DashScope, OpenAI, Kling, and others. It also accepts fixed scripts and custom uploaded footage, and offers visual templates, so a channel can keep a consistent look. Teams already running ComfyUI can drop in their own pipelines, while less technical users can start from the bundled Windows package.