AutoClip is an AI-powered tool for automatically clipping and generating highlight reels from videos, primarily targeting content creators and editors. It simplifies the process of identifying key moments, creating curated collections, and downloading video segments, offering a modern web interface for interaction.

How It Works

The system leverages AI models (DashScope or SiliconFlow) to analyze video content, extract outlines, generate timelines, score segments, and suggest relevant collections. It supports Bilibili video downloads and subtitle extraction, feeding this data into the AI pipeline for intelligent slicing. Users can then manually refine these clips and collections via a drag-and-drop interface.

Quick Start & Requirements

• Installation: Clone the repository, install backend dependencies (pip install -r requirements.txt), and frontend dependencies (npm install).
• Prerequisites: Python 3.8+, Node.js 16+, and an API key for either DashScope or SiliconFlow.
• Configuration: Copy data/settings.example.json to data/settings.json and insert your API key.
• Running: Execute ./start_dev.sh for development, or manually start the backend (python backend_server.py) and frontend (npm run dev). Command-line processing is also available via main.py.
• Access: Frontend at http://localhost:3000, Backend API at http://localhost:8000.
• Docs: API documentation available at http://localhost:8000/docs.

Highlighted Details

• Supports both DashScope (Qwen models) and SiliconFlow (Qwen2.5, DeepSeek-R1) AI providers.
• Features a modern React + TypeScript + Ant Design web interface.
• Offers one-click package download for all generated clips and collections.
• Includes a command-line interface for direct video processing and project management.

Maintenance & Community

The project has recent updates (v1.1.0 as of August 2025) adding SiliconFlow support and improving the LLM client management. Contact information includes QQ, Feishu, and email (christine_zhouye@163.com).

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

Bilibili video download functionality requires the user to be logged into Bilibili, and the correct browser must be selected for the download to succeed. AI analysis speed and clip quality can be tuned by adjusting chunk_size and min_score_threshold parameters, respectively.