Show HN: Reko – Local-First YouTube-to-Markdown LLM Summarizer (github.com)

Reko, a new command-line tool, empowers users to convert YouTube videos into clean Markdown summaries by utilizing transcripts and Large Language Models (LLMs). By simply inputting a video URL, playlist, or multiple URLs, reko fetches the transcript, processes it into concise summaries, and generates accessible documentation. Designed for speed and automation, it optimally handles various video lengths and offers multi-language support, including automatic fallback and translation. The significance of reko lies in its ability to streamline content consumption from lengthy videos, especially valuable for educators, researchers, and anyone seeking to distill essential information quickly. It operates primarily as a local-first application with an emphasis on user privacy but also accommodates cloud model integrations when necessary. Technical features include support for Small Language Models (SLMs), efficient transcript chunking, and the option to extract key points, enhancing usability for diverse summarization tasks. As a versatile tool built on Python, reko represents a notable advancement in making digital content more digestible and user-friendly.