Show HN: AutoShorts – Local, GPU-accelerated AI video pipeline for creators (github.com)

AutoShorts, a new AI-driven tool, has been launched to help content creators easily generate engaging vertical short clips from lengthy gameplay footage. The platform utilizes advanced scene analysis powered by OpenAI's GPT-5-mini or Google Gemini to identify exciting moments, such as action sequences or funny fails, and automatically crops, renders, and adds AI-generated captions or voiceovers. With its GPU-accelerated pipeline, AutoShorts offers features like semantic analysis modes, multilingual support, and customizable video output, making it a versatile tool for creating viral content. This announcement is significant for the AI/ML community as it showcases the potential of AI to streamline video editing processes and enhance content creation workflows. The technical details highlight the use of tools like PyTorch for video rendering and audio analysis via torchaudio, while also integrating local text-to-speech systems to eliminate reliance on cloud APIs. For developers, AutoShorts offers robust configuration options and a straightforward installation process, featuring a Makefile for environment setup. Overall, AutoShorts represents a considerable advancement in automating the creation of shareable video content, empowering creators to effectively engage their audiences with minimal effort.