Pixelle-Video

Pixelle-Video is a fully automated short-video generation engine. Feed it a topic, and it runs the whole pipeline on its own: writing the script, generating the visuals, recording the voiceover and assembling a finished short video ready to publish on social platforms.

Overview

Pixelle-Video has earned more than 29,200 stars on GitHub and comes from Alibaba’s AIDC-AI group, with ongoing development through 2026. Short-form video is where audience attention lives, but producing it consistently is labor-intensive. Pixelle-Video chains the pieces together: a language model drafts the narration, image generation (with ComfyUI workflows supported) produces matching frames, text-to-speech voices the script, and the engine composes everything into a vertical video with timing and transitions handled automatically. Templates and configuration options let you steer style and format, and because the whole stack is self-hosted and open source, creators can run it at volume without per-video platform fees.

Key Features

• End-to-end short-video pipeline: topic in, finished video out
• Automatic script writing, image generation, TTS voiceover and assembly
• ComfyUI integration for flexible, high-quality image workflows
• Configurable templates, voices and styles for different content niches
• Batch generation for channels that publish on a daily schedule
• Self-hosted and open source under the Apache-2.0 license
• Backed by Alibaba’s AIDC-AI team with an active community

💳 Pricing

Pixelle-Video is free and open source under the Apache-2.0 license. Running it costs only the model API usage (LLM, image and TTS services) that you connect, or nothing extra if you use local models.

Who Is It For

Pixelle-Video is for short-video creators, social media marketers, faceless-channel operators and content teams who need a repeatable, automated way to turn ideas into publishable vertical videos at scale.