autoclip is a self-hostable application that automates the extraction and clipping of highlight segments from videos for secondary creation. It uses AI to analyze video content, identify highlights, and generate compilations from sources such as YouTube, Bilibili, or local files.
Project overview
autoclip stands out by providing an end-to-end pipeline that spans from video downloading through AI-powered highlight generation to compilation export, reducing the manual effort required for short-video creation workflows.
Project type
Workflow · Video
Use cases
Video Creation
Deployment
Refer to project documentation
License
MIT
Best for
Content creators and developers who need to automate the extraction of highlight segments from videos for secondary creation.
Users who want AI-driven highlight identification, excitement scoring, and compilation generation within a self-hosted environment.
Key capabilities
Downloads videos from YouTube and Bilibili, and supports local file uploads.
Uses the Qwen LLM to analyze video content, extract outlines, and identify highlight segments.
Automatically cuts videos into highlight segments and scores them based on excitement.
AI recommends and generates video compilations with drag-and-drop ordering support.
Limitations and risks
Mobile support is still under development and is not yet available.
Account management for Bilibili is under development and is not yet available.
The application requires external services including the YouTube API, Bilibili API, and the Qwen LLM API to perform content analysis and video retrieval. AI content analysis depends on a remote model rather than local inference.
Getting started
Clone the repository, configure the .env file, run ./docker-start.sh, and access the Web UI. The setup difficulty is medium due to external dependencies such as FFmpeg and Redis, though scripts are provided to assist with deployment.