PreenCut is an AI-powered tool for automatically clipping video and audio content based on natural language queries. It targets content creators, researchers, and anyone needing to quickly extract specific segments from media, offering efficient retrieval and summarization of spoken content.

How It Works

The tool leverages OpenAI's Whisper for accurate speech-to-text transcription, followed by large language models (LLMs) for content analysis, summarization, and segmentation. Users can query their media using natural language prompts, and PreenCut identifies and extracts relevant clips, which can be exported individually or merged. This approach allows for rapid content discovery and repurposing without manual review.

Quick Start & Requirements

• Install via pip install -r requirements.txt.
• Requires Python 3.9+ and FFmpeg.
• API keys for LLM services (e.g., DeepSeek, DouBao) must be configured in config.py and set as environment variables.
• Gradio interface is launched with python main.py.
• Supports common video and audio formats (mp4, avi, mov, mkv, ts, mxf, mp3, wav, flac).

Highlighted Details

• Automatic Speech Recognition via OpenAI Whisper.
• LLM-powered content analysis and summarization.
• Natural language querying for segment retrieval.
• Smart clipping with options for individual files or merged output.
• SRT subtitle export with timestamp accuracy.
• Batch processing for multiple files.
• Re-analysis feature to experiment with prompts without reprocessing.
• RESTful API for programmatic access.

Maintenance & Community

• Primary contact via email: 1242727205@qq.com.
• Community presence on RedNote (小红书).

Licensing & Compatibility

• Licensed under the MIT License.
• Permissive license suitable for commercial use and integration with closed-source projects.

Limitations & Caveats

Performance can be tuned by adjusting Whisper model size, batch size, and disabling alignment for faster processing if precise timestamps are not critical. The effectiveness of analysis is dependent on the quality of the LLM and the clarity of the user's prompt.